Asperger Syndrome and High Functioning Autism

Asperger Syndrome and High Functioning Autism: Same, Different, or a Spectrum? An Investigation Using a Comprehensive Communication Assessment Battery

Fiona M. Lewis, B Sp Path (Hons), BA School of Health and Rehabilitation Sciences University of Queensland Australia

A thesis submitted for the degree of Doctor of Philosophy at the University of Queensland in July, 2007.

ii

iii

Candidate’s Statement of Originality The material presented in this thesis is the original work of the candidate, except as acknowledged in the text, and has not previously been submitted, either in whole or part, for a degree at this or any other university.

Candidate’s Statement of Contribution of Jointly Published Work All of the jointly authored papers incorporated into this thesis are the original work of the candidate, including all text, tables, and figures.

Statement of Contribution by Others The author of this thesis acknowledges the intellectual input provided by her thesis advisers and paper co-authors, which was to the extent of providing expertise and guidance in the design, implementation, and writing up of the studies reported in the thesis.

I hereby certify that all co-authors have provided the consent for the inclusion of the papers in this thesis and the co-authors accept that the student’s contribution to the paper is as described in the Statement of Originality.

Professor Bruce Murdoch (Principal Supervisor)

Fiona Lewis

iv

v

Acknowledgements I thank all who assisted in the preparation of this thesis. In particular, I extend my gratitude to my supervisors Professor Bruce Murdoch and Dr Gail Woodyatt for their advice, guidance, wisdom, patience, and ongoing encouragement during the course of my research. Without their invaluable input the thesis would not have been possible. Thanks are also extended to Ms Fiona Stroud, IT Support Services, University of Queensland, for her invaluable assistance with the final preparation of the manuscript. I acknowledge with appreciation the assistance provided by the Asperger Syndrome Support Network, Autism Queensland, Education Queensland, Queensland Paediatric Rehabilitation Service, Peggy Frankish (Ipswich Stroke Support Group), Dawn Norris (Gatton Stroke Support Group), and Leon Capra (Principal, St Augustine’s College, Springfield), in the recruitment of participants. Many children, adults, and their families were keen to participate in the research, and welcomed me into their homes. I gratefully acknowledge their interest and support. I gratefully acknowledge the financial assistance from the Zonta Club of Brisbane through the Dr Helen Row-Zonta International Memorial Prize, 2004. Last, but by no means least, I thank my family for their ongoing support, interest, and patience during my postgraduate student days. Arcanum Linguae

vi

vii

List of Prizes, Presentations, and Publications Prizes awarded to the Candidate Relating to the Thesis Dr Helen Row-Zonta International Memorial Prize. (2004). [Research proposal]. Communication Profile of Females with Asperger Syndrome. Presentations by the Candidate Relating to the Thesis Lewis, F. M., Woodyatt, G. C., & Murdoch, B. E. (2006, June). Females with autistic features: Are language and social skills barriers to diagnosis? Paper presented at

the 13th Learning Conference, Jamaica. Publications by the Candidate Relating to the Thesis Lewis, F. M., Murdoch, B. E., & Woodyatt, G. C. (2007). Asperger syndrome and/or high functioning autism: Clinical application of findings from research into language and communication skills. Communicative Disorders Review, 1(1), 3747. Lewis, F. M., Murdoch, B. E., & Woodyatt, G. C. (2007). Linguistic abilities in children with Autism Spectrum Disorder. Research in Autism Spectrum Disorders, 1, 85100. Lewis, F. M., Murdoch, B. E., & Woodyatt, G. C. (2007). Communicative competence and metalinguistic ability: Performance by children and adults with Autism Spectrum Disorder. Journal of Autism and Developmental Disorders, 37, 15251538. Lewis, F. M., Woodyatt, G. C., & Murdoch, B. E. (in press). Linguistic and pragmatic language skills in adults with Autism Spectrum Disorder: A pilot study. Research in Autism Spectrum Disorders.

viii Lewis, F. M., Woodyatt, G. C., & Murdoch, B. E. Females and males with Autism Spectrum Disorder: A comparison study of language and social pragmatic skills. Manuscript submitted for publication. (Autism).

ix

Abstract Language and communication difficulties are central to Asperger Syndrome/Disorder (AS) and Autistic Disorder (AD), two pervasive developmental disorders classified in the Diagnostic and Statistical Manual of Mental Disorders-Fourth Edition (DSM-IV). Research findings from studies investigating language and communication skills in AS and AD may, therefore, have the potential to assist with the development of clinical intervention strategies to promote positive psychosocial, educational, behavioural, and vocational outcomes for children and adults with the disorders. Research into the language and communication skills associated with AS and high-functioning individuals with AD (HFA) reflects the ongoing debate regarding if and how AS fits into the autism paradigm. Three theoretical approaches have been applied to research. These are a) validation studies, where the relevance of developmental language history on linguistic outcomes in AS and HFA has been examined; b) studies where AS and HFA are combined into a single experimental group; and c) studies that view AS and HFA as dimensions on an autism spectrum (ASD). The clinical understanding of the language and communication skills associated with AS and HFA, however, may not be an accurate reflection of the difficulties experienced by individuals with the diagnoses due to three methodological limitations of research to date. Firstly, the extent of language assessment used in studies to date has been restricted. Considerable research has investigated basic linguistic abilities only, such as semantic and syntactic development. This focus reflects DSM-IV’s interest in early semantic and syntactic development as the differential diagnostic criterion between AS and AD. Subsequent development of communicative competence is not a DSM-IV diagnostic issue. Nevertheless, an accurate definition of the communicative difficulties experienced by individuals with a diagnosis of either AS or HFA is a clinical issue if intervention and support are to be specific to the individual’s communicative needs. Secondly, research to date has focused predominantly on children. Long-term studies of adults with developmental histories of language impairment, but not

x necessarily ASD, have demonstrated that communication difficulties may contribute to poor psychosocial and vocational outcomes in adulthood. It has been suggested, moreover, that the language and communication difficulties associated with ASD may contribute to reduced psychosocial outcomes in adults with the disorder. Despite this, there is a dearth of information regarding the communicative skills in this population. Clinically relevant research findings are needed to define the language and communication skills in adults with ASD to determine if intervention and support for adults is warranted. Thirdly, researchers have argued that there may be a female phenotype of autism, where females have better language and social skills. Females with a diagnosis of ASD have been included in many studies, but no study to date has undertaken comparative analyses of the communication skills for the two genders. The absence of clinically useful information regarding the language and social skills in females with ASD convinces of the need for investigation. The overall aim of the thesis was, therefore, to provide clinically relevant findings to extend the understanding of the communicative needs of children, adults, and females with a diagnosis of ASD and no documented intellectual impairment by comprehensively investigating language and communicative abilities that emerge subsequent to the initial onset of basic semantic and syntactic structures. Where possible, the three theoretical perspectives of previous research provided the framework for the investigations. To this end, a group of 20 children (16 males and 4 females), aged 9 years 0 months – 17 years 1 month, and a group of 17 adults (9 females and 8 males), aged 18 years – 67 years with a diagnosis of AS, HFA, or ASD and no documented intellectual impairment were investigated using a range of objective and subjective assessments. Chapters 2, 3, 4, and 5 of this thesis report on investigations into the linguistic, metalinguistic, pragmatic, and social problem solving abilities of the children with the diagnoses. Chapters 6, 7, and 8 focus on these skills in the adults. Findings from the child studies suggest that children reclassified as AS or HFA,

xi based on reported developmental language history, could be differentiated on tasks requiring the resolution of ambiguity and the interpretation of metaphors presented pictorially. A weakness of this approach was the reliance on retrospective recall of developmental language history. When grouped as a single experimental group of ASD participants, the children presented with linguistic deficits in core language, receptive language, expressive language, language content, and language memory. Metalinguistic deficits were identified in resolving ambiguity, understanding inferential language, and in the ability to provide logical, meaningful, and contextually-appropriate sentences. Pragmatic deficits were identified, encompassing the production of emphatic stress to convey meaning in speech and lexical-semantic flexibility. Based on subjective Informant ratings of pragmatic skills, there were significant differences between the ASD group of children and the control group on coherence, inappropriate initiation, stereotypical use of language, use of context, and nonverbal communication skills. When the language effect was removed from the analysis, the ASD group was identified as having deficits in social problem solving relative to the control group of typically developing children. A limitation of combining all participants into a single experimental group was the noted heterogeneity of skills within the non-differentiated experimental group. The range of skills within the ASD group of children was subsequently examined through cluster analysis. Subgroups were described, with performance across the assessments ranging from average performance to severe difficulties. Viewing AS and HFA as disorders on an autism spectrum and examining within-group differences provided clinically useful information for intervention planning. When grouped as a single experimental ASD group, the adults were less proficient than the control group of adults with a typical developmental history on a range of linguistic skills, including overall linguistic ability, auditory verbal comprehension, and naming skills. Pragmatic deficits included difficulties with inferential and figurative language, linguistic flexibility, and the production of variations in emphatic stress. Social problem solving difficulties, when the effect of language skill was removed, were likewise evident in the adults relative to typical peers. Heterogeneity was noted, but not further examined in the combined experimental adult group. An investigation of

xii subgroups of adults within the autism spectrum revealed the heterogeneity of the skills associated with the disorder. Performance of the adults across the assessments ranged from average skills to severe difficulties. Chapter 9 reports on a comparative analysis of language and social pragmatic skills in female and male adults with a diagnosis of ASD. The findings propose that neither language skills nor pragmatic social skills differentiated the females from the males. The overall findings of the thesis suggest that the language and communication skills within the non-intellectually impaired ASD presentation are heterogeneous. Reduced competence may be evident throughout childhood and extend into adulthood. In adulthood, the language and social skills of females are not significantly different to males with the disorder. Based on the series of studies presented in this thesis, conceptualising AS and HFA as presentations on a spectrum of autistic disorders provides the most reliable and clinically applicable descriptions of the language and communication strengths and weaknesses of individuals presenting with the diagnoses. These findings should be considered in intervention planning, and hence, are relevant to speech pathologists, educators, and vocational support workers involved in providing services to the ASD population. General limitations of the study and suggestions for future research are presented in the final chapter.

xiii

Table of Contents Statement of Originality…….….…………………………………………..……………..iii Acknowledgements……..……………………………………………..............................v List of Prizes, Presentations, and Publications………...………………..……………vii Abstract……………………………......…..…………………………...……..………….ix Table of Contents…………..……………………………………………...…..…….…xiii List of Tables………………………...…………………………………..……...….......xxi List of Figures…...…………………………………………………....…….…….…..xxiii List of Abbreviations…………………………….…………….....……………...........xxvi

1

Introduction .............................................................................................................. 1 1.1

Historical Overview .................................................................................... 2

1.2

DSM-IV Diagnostic Criteria for AS and AD .............................................. 3

1.3

Current Research Approaches..................................................................... 4

1.3.1

Validation Studies....................................................................................... 4

1.3.2

Combined AS/HFA Studies........................................................................ 7

1.3.3

Spectrum Studies ........................................................................................ 7

1.4

Methodological Concerns of Research to Date .......................................... 9

1.4.1

Range of Assessments................................................................................. 9

1.4.2

Language and Communication Skills in Adulthood................................. 11

1.4.3

The Female Presentation........................................................................... 12

1.4.4

Influence of Cognition .............................................................................. 13

1.5

Rationale for the Series of Studies............................................................ 13

xiv 1.6 2

Aims of the Thesis .................................................................................... 14

Linguistic Abilities in Children with Autism Spectrum Disorder .................... 16 2.1

Introduction............................................................................................... 16

2.1.1

Validation Studies..................................................................................... 17

2.1.2

Spectrum Approach .................................................................................. 18

2.2

Rationale for the Present Study................................................................. 19

2.3

Aims and Hypotheses of the Present Study .............................................. 19

2.4

Method ...................................................................................................... 20

2.4.1

Participants................................................................................................ 20

2.4.2

Measures ................................................................................................... 24

2.4.3

Procedure .................................................................................................. 25

2.5

Results....................................................................................................... 26

2.5.1

Performance on the CELF-4 by the Children with ASD Reclassified as AS or HFA Based on Reported Developmental Language History ..... 26

2.5.2

Performance by all Child Participants on the CELF-4 ............................. 28

2.5.3

Clusters of Children with ASD Based on Performance on the CELF4................................................................................................................. 35

2.6

Discussion ................................................................................................. 40

2.6.1

Linguistic Skills in Children with ASD Reclassified as AS or HFA According to Reported Developmental Language History....................... 41

2.6.2

Linguistic Skills in Children with ASD .................................................... 43

2.6.3

Heterogeneity of Linguistic Skills in Children with ASD ........................ 44

2.7 3

Conclusion ................................................................................................ 45

Communicative Competence and Metalinguistic Ability in Children with Autism Spectrum Disorder .................................................................................. 48 3.1

Introduction............................................................................................... 48

3.2


3.3


3.4

Method ...................................................................................................... 50

3.4.1

Participants................................................................................................ 50

xv 3.4.2

Measures ................................................................................................... 50

3.4.3

Procedure .................................................................................................. 51

3.5 3.5.1

Results....................................................................................................... 51 Performance on the TLC-E by the Children with ASD Reclassified as AS or HFA Based on Reported Developmental Language History ..... 51

3.5.2

Performance by all Child Participants on the TLC-E ............................... 53

3.5.3

Clusters of Children with ASD Based on Performance on the TLC-E..... 58

3.6 3.6.1

Discussion ................................................................................................. 64 Metalinguistic Abilities in Children with ASD Reclassified as AS or HFA According to Reported Developmental Language History.............. 64

3.6.2

Metalinguistic Language Abilities in Children with ASD........................ 65

3.6.3

Heterogeneity of Metalinguistic Skills in Children with ASD ................. 66

3.7 4

Conclusion ................................................................................................ 68

Pragmatic Language Skills in Children with Autism Spectrum Disorder ........ 69 4.1 4.1.1

Introduction............................................................................................... 69 Validation Approach to Investigating Pragmatic Skills in AS and HFA........................................................................................................... 70

4.1.2

Combined Approach to Investigating Pragmatic Skills in AS and HFA........................................................................................................... 71

4.1.3

AS-specific Studies Investigating Pragmatic Skills.................................. 73

4.1.4

Spectrum Approach to Investigating Pragmatic Skills in ASD ................ 73

4.2


4.3


4.4

Method ...................................................................................................... 75

4.4.1

Participants................................................................................................ 75

4.4.2

Measures ................................................................................................... 76

4.4.3

Procedure .................................................................................................. 77

4.5

Results....................................................................................................... 79

xvi 4.5.1

Examination of Pragmatic Performance by the Children with ASD Reclassified as AS or HFA Based on Reported Developmental Language History...................................................................................... 79

4.5.1.1 Objective pragmatic assessment using the RHLB. ................................... 79 4.5.1.2 Subjective pragmatic assessment using the CCC-2 ratings. ..................... 81 4.5.2

Examination of Pragmatic Performance by all Child Participants ........... 82

4.5.2.1 Objective pragmatic assessment using the RHLB. ................................... 82 4.5.2.2 Subjective pragmatic assessment using the CCC-2 ratings. ..................... 89 4.5.3

Clusters of Children with ASD Based on Pragmatic Measures................ 95

4.5.3.1 Clusters of children with ASD based on performance on the RHLB. ...... 95 4.5.3.2 Clusters of children with ASD based on the CCC-2 ratings................... 100 4.6 4.6.1

Discussion ............................................................................................... 104 Pragmatic Skills in Children with ASD Reclassified as AS or HFA According to Reported Developmental Language History..................... 104

5

4.6.2

Pragmatic Skills in Children with ASD .................................................. 105

4.6.3

Heterogeneity of Pragmatic Skills in Children with ASD ...................... 107

4.7

Specific Limitations of the Study ........................................................... 110

4.8

Conclusion .............................................................................................. 111

Social Problem Solving in Children with Autism Spectrum Disorder............. 113 5.1

Introduction............................................................................................. 113

5.2

Studies Investigating Social Problem Solving in Children with ASD.... 113

5.2.1

AS-specific Studies ................................................................................. 114

5.3

Rationale for the Present Study............................................................... 116

5.4

Aims and Hypotheses of the Present Study ............................................ 117

5.5

Method .................................................................................................... 118

5.5.1

Participants.............................................................................................. 118

5.5.2

Measures ................................................................................................. 118

5.5.3

Procedure ................................................................................................ 119

5.6

Results..................................................................................................... 120

xvii 5.6.1

Performance on the TOPS by the Children with ASD Reclassified as AS or HFA Based on Reported Developmental Language History ....... 120

5.6.2 5.7 5.7.1

Performance by all Child Participants on the TOPS............................... 121 Discussion ............................................................................................... 123 Social Problem Solving in Children with ASD Reclassified as AS or HFA According to Reported Developmental Language History............ 123

6

5.7.2

Social Problem Solving in Children with ASD ...................................... 124

5.7.3

Clinical Application of the Findings....................................................... 125

5.8

Specific Limitations and Future Research Directions ............................ 127

5.9

Conclusion .............................................................................................. 127

Adults with Autism Spectrum Disorder: Linguistic and Pragmatic Skills .... 129 6.1

Introduction............................................................................................. 129

6.2


6.3


6.4

Method .................................................................................................... 133

6.4.1

Participants.............................................................................................. 133

6.4.2

Measures ................................................................................................. 137

6.4.3

Procedure ................................................................................................ 139

6.5 6.5.1

Results..................................................................................................... 140 Performance by the Adults on the WAB, the RSPCS-SR, the RSPCS-SO, and the RHLB..................................................................... 140

6.5.2

Clusters of Adults with ASD Based on Performance on the WAB AQ and RHLB ........................................................................................ 152

6.6

Discussion ............................................................................................... 156

6.6.1

Linguistic and Pragmatic Skills in Adults with ASD ............................. 156

6.6.2

Heterogeneity of Linguistic and Pragmatic Language Skills of Adults with ASD..................................................................................... 158

7

6.7


6.8

Conclusion .............................................................................................. 160

Metalinguistic Abilities in Adults with Autism Spectrum Disorder ............... 162

xviii 7.1

Introduction............................................................................................. 162

7.2


7.3


7.4

Methods................................................................................................... 164

7.4.1

Participants.............................................................................................. 164

7.4.2

Measures ................................................................................................. 164

7.4.3

Procedure ................................................................................................ 165

7.5 7.5.1

Performance by the Adults on the TLC-E .............................................. 165

7.5.2

Clusters of Adults with ASD Based on Performance on the TLC-E...... 170

7.6

Discussion ............................................................................................... 174

7.6.1

Metalinguistic Abilities in Adults with ASD.......................................... 174

7.6.2

Heterogeneity of Metalinguistic Skills in Adults with ASD .................. 175

7.7 8

Results..................................................................................................... 165

Conclusion .............................................................................................. 176

Social Problem Solving in Adults with Autism Spectrum Disorder................. 178 8.1

Introduction............................................................................................. 178

8.2

Verbal and Social Problem Solving in Adults with ASD ....................... 178

8.3


8.4


8.5

Method .................................................................................................... 182

8.5.1

Participants.............................................................................................. 182

8.5.2

Measures ................................................................................................. 182

8.5.3

Procedure ................................................................................................ 182

8.6

Results..................................................................................................... 183

8.7

Discussion ............................................................................................... 185

8.7.1

Social Problem Solving in Adults with ASD.......................................... 185

8.7.2

Heterogeneity of Social Problem Solving Skills in Adults with ASD.... 185

8.7.3

Clinical Utility of the Findings ............................................................... 187

8.8

Specific Limitations and Future Research Directions ............................ 189

8.9

Conclusion .............................................................................................. 190

xix 9

Females and Males with Autism Spectrum Disorder: A Comparison Study of Language and Social Pragmatic Skills .......................................................... 191 9.1

Introduction............................................................................................. 191

9.2


9.3

Aim and Hypothesis of the Present Study .............................................. 193

9.4

Method .................................................................................................... 193

9.4.1

Participants.............................................................................................. 193

9.4.2

Measures ................................................................................................. 195

9.4.3

Procedure ................................................................................................ 195

9.5

Results..................................................................................................... 195

9.6

Discussion ............................................................................................... 202

9.6.1

Language Skills in Females with ASD ................................................... 203

9.6.2

Social Skills in Females with ASD......................................................... 203

9.7


9.8

Conclusion .............................................................................................. 205

10

Conclusions............................................................................................................ 206

10.1

Summary and Conclusions ..................................................................... 206

10.1.1

Child Studies........................................................................................... 206

10.1.2

Adult Studies........................................................................................... 208

10.1.3

Gender Study .......................................................................................... 210

10.2

Limitations of the Present Research ....................................................... 211

10.3

Future Research Directions..................................................................... 215

References ...................................................................................................................... 218 Appendixes..................................................................................................................... 242

Appendix A DSM-IV diagnostic criteria for Autistic Disorder .................................... 242 Appendix B DSM-IV diagnostic criteria for Asperger Disorder/syndrome ................. 244 Appendix C Portion of questionnaire pertaining to developmental language history....................................................................................................... 246

xx Appendix D Examples of the four Test of Language Competence-Expanded Edition subtests ........................................................................................ 247 Appendix E Examples of subtest items from the Right Hemisphere Language Battery...................................................................................................... 250 Appendix F Example items from the Children’s Communication ChecklistSecond Edition ......................................................................................... 255 Appendix G Example scenario, related questions, and scoring guide for Picture 1 of the Test of Problem Solving-Elementary, Revised.............................. 256 Appendix H Example script, related questions, and scoring guide for the Test of Problem Solving-Adolescent ................................................................... 258 Appendix I

Rating Scale of Pragmatic Communication Skills................................... 260

Appendix J

Assessments administered to all adult participants in the gender study ......................................................................................................... 262

xxi

List of Tables Table 2.1

ASD child participant demographics ......................................................... 22

Table 2.2

Descriptive statistics of the performance on the CELF-4 by children with ASD reclassified as AS or HFA according to developmental language history ......................................................................................... 27

Table 2.3

Descriptive statistics of the child participants on the CELF-4................... 28

Table 2.4

Subgroups of ASD children as determined by the CELF-4 Index Scores showing standard deviations from the control group means .......... 38

Table 3.1

Descriptive statistics of the performance on the TLC-E by children with ASD reclassified as AS or HFA according to developmental language history ......................................................................................... 52

Table 3.2

Descriptive statistics of the child participants on the TLC-E .................... 53

Table 3.3

Subgroups of ASD children based on performance on the TLC-E............ 61

Table 4.1

Pearson correlations between ratings by Informants 1 and Informants 2 on the 10 CCC-2 scales for the 35 children.......................... 78

Table 4.2

Descriptive statistics of the performance on the RHLB by children with ASD reclassified as AS or HFA according to developmental language history ......................................................................................... 80

Table 4.3

Descriptive statistics of Informants Number 1 ratings on the CCC-2 for children with ASD reclassified as AS or HFA according to developmental language history................................................................. 81

Table 4.4

Descriptive statistics of the child participants on the RHLB ..................... 83

Table 4.5

Descriptive statistics of the child participants on the CCC-2 scales based on ratings by Informants Number 1 ................................................. 90

Table 4.6

Subgroups of ASD children based on performance on the RHLB ............ 98

Table 4.7

Subgroups of ASD children based on Informants Number 1 ratings on the CCC-2 ........................................................................................... 102

Table 5.1

Descriptive statistics of the performance on the TOPS by children with ASD reclassified AS and HFA according to developmental language history ....................................................................................... 121

Table 5.2

Means, standard deviations, and ranges of scores for the child participant groups on Core Language (CELF-4) and the TOPS.............. 122

Table 6.1

ASD adult participant demographics ....................................................... 134

xxii Table 6.2

Descriptive statistics of the performance on the WAB, the RHLB, the RSPCS-SR, and RSPCS-SO by the adult participants....................... 142

Table 6.3

Subgroups of adults with ASD based on performance on the WAB AQ and the six RHLB subtests ................................................................ 154

Table 7.1

Descriptive statistics of the adult participants on the TLC-E .................. 166

Table 7.2

Subgroups of ASD adults based on performance on the TLC-E ............. 172

Table 8.1

Means, standard deviations, and ranges of performance scores from the adult participants on the verbal and problem solving measures ........ 183

Table 9.1

Female and male ASD adult participant demographics........................... 194

Table 9.2

Descriptive statistics of the four adult comparative groups ..................... 196

Table 9.3

Language and social skill differences between ASD females, ASD males, control females, and control males ............................................... 199

Table 9.4

Correlations between the Significant Other rating on the RSPCS and measures of language performance for the four comparative adult groups.............................................................................................. 200

xxiii

List of Figures Figure 2.1.

Box plots displaying the dispersal of scores from the child ASD and control groups’ performance on the Core Language Index from the Clinical Evaluation of Language Fundamentals-4th Edition...................... 30

Figure 2.2.

Box plots displaying the dispersal of scores from the child ASD and control groups’ performance on the Receptive Language Index from the Clinical Evaluation of Language Fundamentals-4th Edition........................................................................................................ 31

Figure 2.3.

Box plots displaying the dispersal of scores from the child ASD and control groups’ performance on the Expressive Language Index from the Clinical Evaluation of Language Fundamentals-4th Edition........................................................................................................ 32

Figure 2.4.

Box plots displaying the dispersal of scores from the child ASD and control groups’ performance on the Language Content Index from the Clinical Evaluation of Language Fundamentals-4th Edition. ............... 33

Figure 2.5.

Box plots displaying the dispersal of scores from the child ASD and control groups’ performance on the Language Memory Index from the Clinical Evaluation of Language Fundamentals-4th Edition. ............... 34

Figure 2.6.

Agglomerative hierarchical cluster analysis of the 20 ASD children based on performance on the five Index Scores from the Clinical Evaluation of Language Fundamentals-4th Edition.................................... 36

Figure 3.1.

Box plots displaying the dispersal of scores for the child ASD and control groups’ performance on Ambiguous Sentences. ........................... 55

Figure 3.2.

Box plots displaying the dispersal of scores for the child ASD and control groups’ performance on Listening Comprehension: Making Inferences. .................................................................................................. 56

Figure 3.3.

Box plots displaying the dispersal of scores for the child ASD and control groups’ performance on Oral Expression: Recreating Sentences.................................................................................................... 57

Figure 3.4.

Box plots displaying the dispersal of scores for the child ASD and control groups’ performance on Figurative Language............................... 58

Figure 3.5.

Agglomerative hierarchical cluster analysis of the 20 ASD children based on performance on the four subtests from the Test of Language Competence-Expanded Edition................................................. 59

Figure 4.1.

Box plots displaying the dispersal of scores from the child ASD and control groups’ performance on Metaphor Picture Test. ........................... 84

xxiv Figure 4.2.

Box plots displaying the dispersal of scores from the child ASD and control groups’ performance on Written Metaphor Test. .......................... 85

Figure 4.3.

Box plots displaying the dispersal of scores from the child ASD and control groups’ performance on Comprehension of Inferred Meaning. .................................................................................................... 86

Figure 4.4.

Box plots displaying the dispersal of scores from the child ASD and control groups’ performance on Appreciation of Humour. ....................... 87

Figure 4.5.

Box plots displaying the dispersal of scores from the child ASD and control groups’ performance on Lexical Semantic Test. ........................... 88

Figure 4.6.

Box plots displaying the dispersal of scores from the child ASD and control groups’ performance on Production of Emphatic Stress. .............. 89

Figure 4.7.

Box plots displaying the dispersal of scores from the child ASD and control groups’ performance on Informants Number 1 rating for Coherence................................................................................................... 91

Figure 4.8.

Box plots displaying the dispersal of scores from the child ASD and control groups’ performance on Informants Number 1 rating for Inappropriate Initiation. ............................................................................. 92

Figure 4.9.

Box plots displaying the dispersal of scores from the child ASD and control groups’ performance on Informants Number 1 rating for Use of Context. .......................................................................................... 93

Figure 4.10. Box plots displaying the dispersal of scores from the child ASD and control groups’ performance on Informants Number 1 rating for Stereotyped Language................................................................................ 94 Figure 4.11. Box plots displaying the dispersal of scores from the child ASD and control groups’ performance on Informants Number 1 rating for Nonverbal Communication. ....................................................................... 95 Figure 4.12. Agglomerative hierarchical cluster analysis of the 20 ASD children based on performance on the six subtests from the Right Hemisphere Language Battery................................................................... 96 Figure 4.13. Agglomerative hierarchical cluster analysis of the 19 ASD children based on Informants Number 1 ratings on the five scales from the Children’s Communication Checklist-Second Edition. ........................... 101 Figure 5.1.

Box plots displaying the dispersal of scores for the child ASD and control groups’ performance on the tests of problem solving (Test of Problem Solving-Elementary, Revised or the Test of Problem Solving-Adolescent)................................................................................. 123

Figure 6.1.

Box plots displaying the dispersal of scores from the adult ASD and control groups’ performance on the Western Aphasia Battery Aphasia Quotient...................................................................................... 144

xxv Figure 6.2.

Box plots displaying the dispersal of scores from the adult ASD and control groups’ performance on the Metaphor Picture Test. ................... 145

Figure 6.3.

Box plots displaying the dispersal of scores from the adult ASD and control groups’ performance on the Written Metaphor Test. .................. 146

Figure 6.4.

Box plots displaying the dispersal of scores from the adult ASD and control groups’ performance on Comprehension of Inferred Meaning. .................................................................................................. 147

Figure 6.5.

Box plots displaying the dispersal of scores from the adult ASD and control groups’ performance on Appreciation of Humour. ..................... 148

Figure 6.6.

Box plots displaying the dispersal of scores from the adult ASD and control groups’ performance on the Lexical Semantic Test. ................... 149

Figure 6.7.

Box plots displaying the dispersal of scores from the adult ASD and control groups’ performance on Production of Emphatic Stress. ............ 150

Figure 6.8.

Box plots displaying the dispersal of scores from the adult ASD and control groups’ performance on the Rating Scale of Pragmatic Communication Skills-Self Rating. ......................................................... 151

Figure 6.9.

Box plots displaying the dispersal of scores from the adult ASD and control groups’ performance on the Rating Scale of Pragmatic Communication Skills-Significant Other Rating. .................................... 152

Figure 6.10. Agglomerative hierarchical cluster analysis of the 17 ASD adults based on performance on the Western Aphasia Battery Aphasia Quotient and the six subtests from the Right Hemisphere Language Battery. ..................................................................................................... 153 Figure 7.1.

Box plots displaying the dispersal of scores from the adult ASD and control groups’ performance on Ambiguous Sentences. ......................... 167

Figure 7.2.

Box plots displaying the dispersal of scores from the adult ASD and control groups’ performance on Listening Comprehension: Making Inferences. ................................................................................................ 168

Figure 7.3.

Box plots displaying the dispersal of scores from the adult ASD and control groups’ performance on Oral Expression: Recreating Sentences.................................................................................................. 169

Figure 7.4.

Box plots displaying the dispersal of scores from the adult ASD and control groups’ performance on Figurative Language............................. 170

Figure 7.5.

Agglomerative hierarchical cluster analysis of the 17 ASD adults based on performance on the four subtests from the Test of Language Competence-Expanded Edition............................................... 171

Figure 8.1.

Box plots displaying the dispersal of scores from the adult ASD and control groups’ performance on the Test of Problem SolvingAdolescent................................................................................................ 184

xxvi

List of Abbreviations AD = Autistic Disorder ADI-R = Autism Diagnostic Interview-Revised ADOS-G = Autism Diagnostic Observation Schedule-Generic Amb Sent = Ambiguous Sentences subtest from the TLC-E A H = Appreciation of Humour test from the RHLB AS = Asperger syndrome/Asperger Disorder ASD = Autism Spectrum Disorder CCC-2 = Children’s Communication Checklist-Second Edition CELF-4 = Clinical Evaluation of Language Fundamentals-Fourth Edition C I M = Comprehension of Inferred Meaning test from the RHLB Core Lge = Core Language Index score from the CELF-4 DSM-IV = Diagnostic and Statistical Manual of Mental Disorders-Fourth Edition Exp Lge = Expressive Language Index score from the CELF-4 Fig Lge = Figurative Language subtest from the TLC-E HFA = High functioning autism ICD-10 = International Classification of Mental and Behavioural Disorders (10th edition). Lge Cont = Language Content Index score from the CELF-4 Lge Mem = Language Memory Index score from the CELF-4 List Comp = Listening Comprehension: Making Inferences subtest from the TLC-E L S = Lexical Semantic Test from the RHLB M P = Metaphor Picture Test from the RHLB Oral Exp = Oral Expression: Recreating Sentences subtest from the TLC-E P E S = Production of Emphatic Stress test from the RHLB Rec Lge = Receptive Language Index score from the CELF-4 RHLB = Right Hemisphere Language Battery RSPCS = Rating Scale of Pragmatic Communication Skills

xxvii RSPCS-SO = Rating by a Significant Other on the Rating Scale of Pragmatic Communication Skills RSPCS-SR = Self-rating on the Rating Scale of Pragmatic Communication Skills TLC-E = Test of Language Competence-Expanded Edition TONI-2 = Test of Nonverbal Intelligence-Second Edition TOPS-A = Test of Problem Solving-Adolescent TOPS-E, R = Test of Problem Solving-Elementary, Revised WAB = Western Aphasia Battery WAB AQ = Western Aphasia Battery Aphasia Quotient W M = Written Metaphor Test from the RHLB

1

1

Introduction 1

Impairments in language and communication skills are integral aspects of the presentations of both Asperger syndrome (AS) and high functioning autism (HFA). Definitive descriptions of the associated language and communicative profiles would, therefore, have clinical relevance in the long-term provision of intervention and support. The aim of this review of the literature to date is to determine the applicability of previous research findings to increasing the clinical understanding of language and communication skills in AS and HFA. The current literature review identifies three differing foci of AS/HFA research. The theoretical framework for each of the three research approaches reflects the ongoing debate regarding if and how AS fits into the autism paradigm. The three theoretical approaches to research are validation studies (e.g., Howlin, 2003; Mayes & Calhoun, 2001), combined AS/HFA studies (e.g., Booth, Charlton, Hughes, & Happe, 2003; Emerich, Creaghead, Grether, Murray, & Grasha, 2003; Losh & Capps, 2003; Ozonoff & Miller, 1996), and studies that examine the range of skills on a spectrum of autistic disorders (Prior et al., 1998; Verte, Geurts, Roeyers, Rosseel, et al., 2006). Historically, the term AS has been linked with autism (Baron-Cohen, Jolliffe, Mortimore, & Robertson, 1997; Klin & Volkmar, 2003; Schopler, 1985; Tantam, 2003). AS has been referred to as higher level autism (Schopler), high functioning autism (Baron-Cohen et al., Mesibov, Shea, & Adams, 2001), little more than high-IQ autism (Miller & Ozonoff, 2000), a milder version of high functioning autism (Miller &

1

“Asperger syndrome and/or high functioning autism: Clinical application of findings from research into

language and communication skills,” by F. M. Lewis, B. E. Murdoch, and G. C. Woodyatt, 2007, Communicative Disorders Review, 1(1) 37-47, was adapted from this chapter.

2 Ozonoff), falling within the autism spectrum disorders (Wing, 1981), a mild variant of autism (Gillberg, 1989), a mild form of autism (Stone, Baron-Cohen, & Knight, 1998), able autistic people (Frith, 1991), and autism without mental retardation (Croen, Grether, Hoogstrate, & Selvin, 2002). The interchangeability of the terms reflect the historical background of AS and its relationship to Autistic Disorder (AD) which provides the theoretical perspective for the three different research approaches. 1.1

Historical Overview

In 1943, Leo Kanner, an Austrian-born American child psychiatrist published Autistic Disturbances of Affective Contact (Kanner, 1942-3). The following year Hans Asperger, an Austrian-born German paediatrician, published a thesis titled Autistic Psychopathy (Frith, 1991). Unaware of each other’s work, they independently described similar presentations. Both wrote of children with disturbing deficits, and both used the term autistic, a term initially used by the Swiss psychiatrist Bleuler in 1916 (Frith, 1991). Bleuler’s use of the word was to describe a schizophrenic patient’s loss of contact with the world. The choice of word relates to its Greek origins (auto = self). Asperger (1991) described “the autist is only himself” (p. 38). Kanner (1942-3) provided detailed descriptions of eight male and three female children who displayed a marked limitation in spontaneity, stereotypic movements, no eye contact, rigidity of behaviour, no awareness of people around them, limited interaction with the environment, complete absorption in things, limited affective contact, and good motor coordination and manual dexterity. According to Kanner, the fundamental presentation was the “inability to relate themselves in the ordinary way to people and situations from the beginning of life” (p. 242). Kanner’s (1942-3) description of the communicative skills of these children suggests delays in early language acquisition. The children had shown early skills of repetition of words, but normal development of language was absent. For instance, the communication skills of these children were described by Kanner as “semantically and conversationally valueless. . . . there is no fundamental difference between the eight

3 speaking and the three mute children. . . . When sentences are finally formed, they are for a long time mostly parrot-like repetitions of heard word combinations” (p. 243). Meanwhile, Asperger (1991) described the characteristics of four boys seen in his clinic. His description included extremely limited relationships with the outside world, stereotypic movements, not reacting to stimuli appropriately, following their own internally generated responses, inappropriate responses to people, things, and situations, rigidity of behaviour, unfocused eye gaze, difficulty learning practical skills, deep absorption in their own preoccupations, and clumsiness. Asperger (1991) reported on the boys’ difficulties of social integration. Characteristics noted were speech and language onset usually being at the appropriate developmental time, peculiarity of eye gaze, poor use and comprehension of facial expression and gesture, an absence of a sense of humour, unnatural or original use of language, a narrow field of interest, difficulties with the mechanics of learning, restricted social relationships, a range of cognitive abilities, and a disharmony between emotion and disposition. Asperger claimed the “fundamental disorder of autistic individuals is the limitation of their social relationships” (p. 77). The accounts by Kanner (1942-3) and Asperger (1991) both included deficits in social communication. However, language onset and development appeared to differentiate Kanner’s Infantile Autism from Asperger’s Autistic Psychopathy. Kanner’s children were characterised by their delayed onset of language and the use of echolalia. In contrast, Asperger’s children were described as having normal onset and development of language, but with subsequent novel or original use of language. 1.2

DSM-IV Diagnostic Criteria for AS and AD

Prior to 1994, AS was listed in the Diagnostic and Statistical Manual of Mental Disorders (DSM) as one of a number of pervasive developmental disorders, where apart from AD, no generally recognised subtypes were described (American Psychiatric Association [APA], 1987). In 1994, however, AS was listed as a categorically distinct

4 presentation (APA, 1994). AS and AD are currently two of five pervasive developmental disorders listed separately in DSM-IV. The 1994 diagnostic distinction between AS and AD challenged the perceived historical relationship between AS and AD. Reflecting the differences in language onset described by Kanner (1942-3) and Asperger (1991), developmental language history is the pertinent diagnostic distinction between AS and AD in DSM-IV (APA, 1994). For a diagnosis of AS, DSM-IV stipulates that while nonverbal communication skills may be impaired, there is to be “no clinically significant general delay in language (e.g., single words used by age 2 years, communicative phrases used by age 3 years)” (p. 77). In contrast, a diagnosis of AD (which includes HFA) requires a “delay in, or total lack of, the development of spoken language” (p. 70). AS and AD both present with the triad of autistic symptoms (impaired social interaction and relationships, abnormal communication, and rigid and limited imagination and play) (Bachevalier, 1994; Bradshaw, 2001; Maurer & Damasio, 1982; Rutter, 1978; Tantam, 1991; Wing, 1989), again reflecting the original descriptions (Asperger, 1991; Kanner, 1942-3). Complete DSM-IV (APA, 1994) criteria for AD and AS are presented in Appendixes A and B respectively. 1.3

Current Research Approaches

1.3.1

Validation Studies

The establishment of the categorical distinction between AS and AD in DSM-IV (APA, 1994) has resulted in ongoing debate regarding the external validity of the diagnosis of AS from the diagnosis of AD with average intelligence (HFA) (e.g., Frith, 2003; Klin, Pauls, Schultz, & Volkmar, 2005; Klin & Volkmar, 2003; Leekam, Libby, Wing, Gould, & , 2000; Macintosh & Dissanayake, 2004; Mayes, Calhoun, & Crites, 2001; Miller & Ozonoff, 1997, 2000; Rutter, 1978; Szatmari, Bartolucci, Finlayson, & Krames, 1986). The theoretical basis for this research focus reflects the historical dilemma of determining the relationship, if any, between AS and AD. In attempts to determine the external

5 validation of AS as a separate diagnosis, a number of studies have examined the relevance of developmental language history on linguistic outcomes in AS and HFA (Ghaziuddin et al., 2000; Howlin, 2003; Mayes & Calhoun, 2001; Szatmari, Archer, Fisman, Streiner, & Wilson, 1995; Szatmari, Tuff, Finlayson, & Bartolucci, 1990). Although there are broader aspects of language development such as morphological and syntactic content, meaning, and structure (Semel, Wiig, & Secord, 2003), the studies by Howlin, Mayes and Calhoun, Ghaziuddin and colleagues, and Szatmari and coresearchers have been restricted to basic semantic skills only, as subsequent development of communicative competence is not a DSM-IV diagnostic issue. Mayes and Calhoun (2001) examined whether a history of early speech/language delay with the attendant diagnosis of autism, and a history of normal language onset and development and the associated diagnosis of AS, had significantly different outcomes for children. Using data from parent interviews, clinical observations of the child, teacher reports, and previous evaluations to complete The Checklist for Autism in Young Children (Mayes & Calhoun, 1999), they determined that there were no significant differences on any of the expressive language measures between the two groups. With these results, Mayes and Calhoun (2001) argued against the validity of AS as distinct from HFA, and called for the removal of AS from the next DSM. In a subsequent work, Mayes and Calhoun (2003) argued that “early speech delay may be irrelevant to later childhood outcome in children who have ASD [autism spectrum disorder] and normal intelligence, and that the absence of a speech delay as a DSM-IV criterion for Asperger’s disorder (vs. autism) may not be justified” (p. 21). Just as Mayes and Calhoun (2001) determined an irrelevance of language history in the outcome of young autistic children, another study, involving adults, supported their conclusions. Howlin (2003) examined the outcome of adults with autism, with and without a history of early language delay, and, using the British Picture Vocabulary Scale (Dunn, Dunn, Whetton, & Burley, 1997; Dunn, Dunn, Whetton, & Pintillie, 1982) and the Expressive One Word Picture Vocabulary Test (Gardner, 1982), found only marginally significant differences between the two groups on receptive and expressive

6 language skills. In contrast, Szatmari et al. (1995), using the Verbal Comprehension Scale A of the Reynell Developmental Scales (Reynell & Huntley, 1987), the Grammatical Completion Test of the Test of Language Development-2 (Newcomer & Hammill, 1988), and the Word Knowledge, Part 2, of the Oral Vocabulary section of the McCarthy Scales of Children’s Abilities (McCarthy, 1972), described children with HFA as being consistently two standard deviations below the mean on a range of language tests, while the AS group performed just below or within one standard deviation from the mean. The authors concluded differences between AS and HFA may be quantitative rather than qualitative. That is, both are part of the same autistic presentation, but differentiation can be made by the degree of their disability. A small number of studies have examined aspects of communication other than basic linguistic skills to determine the external validity of AS as a separate diagnosis to HFA. Ramberg and colleagues (Ramberg, Ehlers, Nyden, Johansson, & Gillberg, 1996), for instance, argued pragmatic skills do not differentiate AS from HFA. Yet others (Gilchrist et al., 2001) have suggested that of the two disorders, individuals with AS are better conversationalists, with the failure to build on previous exchanges resulting in a difficulty in building reciprocal conversation being evident in HFA (Fine, Bartolucci, Szatmari, & Ginsberg, 1994). The discourse skills of individuals with AS have also been described as more complex than those with HFA (Fine et al.), as individuals with AS use more bridging devices than HFA, but at the same time, are more likely to make errors in the use of cohesive links (Fine et al.). Individuals with AS show a tendency for obsessive and repetitive topic expression (Shriberg et al., 2001), and as a group, score higher than HFA on total pedantic speech scores (Ghaziuddin & Gerstein, 1996). Moreover, significant differences have been identified between the speech and prosodic characteristics of the two disorders (Shriberg et al.), with AS using more functionally useful intonation and less non-useful intonation patterns than HFA (Fine, Bartolucci, Ginsberg, & Szatmari, 1991). In conclusion, investigations of semantic skills of children and adults, and

7 pragmatic language skills of children have been undertaken to determine the validation of AS from HFA. To date, the findings have been inconsistent. 1.3.2

Combined AS/HFA Studies

The second research focus is based on the theoretical assumption that AS and HFA are similar presentations of a single disorder. A number of studies examining language and communication skills have been undertaken where participants have been AS and/or HFA, with no delineation attempted between the two groups. Comparison groups have included normally developing peers and language-impaired controls, matched on measures including age, sex, IQ, socioeconomic status, handedness, and educational level. Compared to normally developing peers and/or language impaired controls, individuals with AS/HFA have been described as experiencing linguistic deficits (e.g., Shields, Varley, Broks, & Simpson, 1996), as well as difficulties with pragmatics and discourse (Booth et al., 2003; Emerich et al., 2003; Losh & Capps, 2003; Ozonoff & Miller, 1996; Rumsey & Hamburger, 1988; Shields et al.). In addition, difficulties with planning and problem solving (Booth et al.; Prior & Hoffmann, 1990; Rumsey, 1985; Rumsey & Hamburger) have been identified. Heterogeneity within the AS/HFA samples has been documented (Rumsey & Hamburger), but, to date, not further investigated. In conclusion, a number of studies, where participants with diagnoses of either AS and/or HFA are combined as a single experimental group, have identified linguistic, pragmatic, and problem solving deficits in children and adults with AS/HFA. Heterogeneity within the sample populations has been noted, possibly reflecting different diagnostic groups within the combined group of participants under investigation (Klin, 2000). 1.3.3

Spectrum Studies

A third theoretical perspective, again reflecting the historical link between AS and autism, is evident in the literature. A recent shift in the terminology used in research, which acknowledges both the similarities and differences in the AS and HFA

8 presentation, is evident. The term autistic continuum/autistic spectrum (ASD) was introduced by Wing (1989) as a means of acknowledging all individuals presenting with the triad of autistic characteristics, irrespective of the severity of the symptoms. It is now being used to describe research populations which have previously been referred to as AS or HFA (e.g., Mandell, Walrath, Manteuffel, Sgro, & Pinto-Martin, 2005; Paul, Augustyn, Klin, & Volkmar, 2005; Prior et al., 1998; Seltzer et al., 2003; Sperry & Mesibov, 2005; Volden, 2004; Warreyn, Roeyers, & De Groote, 2005), even though a diagnosis of ASD does not exist in the International Classification of Diseases-10th edition [ICD-10] (World Health Organization [WHO], 1992), or DSM-IV (APA, 1994). Volden (2004) for instance, adopted the nomenclature of ASD to focus on conversational repair skills in children with diagnoses of autism or pervasive developmental disorder-not otherwise specified. She described the significant likelihood of children with ASD responding to requests for clarification with inappropriate responses when compared to language-matched controls. Similarly, Paul et al. (2005) adopted the ASD nomenclature in their study of expressive and receptive prosodic skills in a group of participants with ASD. They found significant differences between the ASD group and a typically developing control group. While conceptualising AS and HFA as disorders on an autism spectrum, the studies by Volden and Paul and colleagues are, in essence, similar to the combined studies. Both research groups failed to accurately apply the theoretical perspective of the spectrum studies; that is, viewing ASD as a continuum, with a range of skills represented within that continuum, as suggested by Wing (1989). Prior et al. (1998), however, accurately applied the theoretical perspective of a continuum of autistic disorders in their investigation of current functioning of a group of children with a diagnosis of either AS, HFA, or a related pervasive developmental disorder. Measures included social interaction, communication, imagination, chosen selfbehaviour, first- and second-order theory of mind tasks, and verbal abilities. Unlike Volden (2004) and Paul and co-researchers (2005), Prior and colleagues examined the ASD group for differences within the combined group of children. Although all the children had intelligence close to normal range, Prior et al. described three clusters of

9 children within the participant group. The clusters differed significantly on verbal abilities as measured by the Peabody Picture Vocabulary Test-Revised (Dunn & Dunn, 1981) or the British Picture Vocabulary Test (Dunn et al., 1982), and theory of mind tasks. Interestingly, given the focus of the validation studies, developmental language history was not significant in determining the differentiation of the clusters, and Prior et al. cautioned against using developmental history for differential diagnosis. In conclusion, a number of studies adopting the nomenclature of ASD have described semantic, pragmatic, and prosodic deficits in children with the disorder. The methodologies of few spectrum studies, however, have reflected Wing’s (1989) vision of a range of skills represented within the autism continuum. 1.4

Methodological Concerns of Research to Date

All three theoretical approaches reviewed above have the potential to further the clinical understanding of the language and communication skills associated with AS and HFA. Based on the above conclusions, however, there are three methodological concerns that may limit the clinical relevance of research findings to date. This, in turn, may lead to an inadequate provision of intervention services to the AS/HFA population. 1.4.1

Range of Assessments

The majority of the validation studies (Ghaziuddin et al., 2000; Howlin, 2003; Mayes & Calhoun, 2001; Szatmari et al., 1995; Szatmari et al., 1990) have focused on linguistic outcomes, with semantic and grammatical skills being the focus of research. This interest in linguistic skills reflects DSM-IV’s (APA, 1994) focus on early language development. Other validation studies, however, have examined skills beyond basic linguistic ability, by focusing on pragmatic skills in conversation (Fine et al., 1991; Fine et al., 1994; Ghaziuddin & Gerstein, 1996; Gilchrist et al., 2001; Ramberg et al., 1996; Shriberg et al., 2001) Combined studies have likewise examined skills beyond basic linguistic ability by focusing on skills that emerge subsequent to the onset of language in early childhood

10 such as humour (Emerich et al., 2003; Ozonoff & Miller, 1996), inference, indirect requests (Ozonoff & Miller), and problem solving (Rumsey & Hamburger, 1988). Spectrum studies have investigated both early and subsequent language skills. While Prior et al. (1998), for instance, administered an assessment of semantic development to define verbal skills in their group of children with ASD, the studies by Volden (2004) and Paul et al. (2005) examined pragmatic aspects of language use. Studies where language and communication assessments are restricted to basic linguistic skills only may have diagnostic relevance but may offer little in terms of clinically applicable findings. In contrast, studies that include language skills that develop subsequent to the onset of language such as the understanding of homonymy (Doherty, 2000), making word associations (Cronin, 2002; Ervin, 1961; Palermo, 1971), speaking and listening skills (Lloyd, Mann, & Peers, 1998), understanding referentially ambiguous pronouns (Sekerina, Stromswold, & Hestvik, 2004) and complex syntactic structures (Ninio, 2004), production and comprehension of functional contrasts through varied intonation (Wells, Peppe, & Goulandris, 2004), and the ability to produce figurative language (Levorato & Cacciari, 2002) may offer clinically relevant findings that assist with the formulation of intervention services. For the three theoretical approaches to provide clinically applicable findings, it may be necessary for each approach to comprehensively assess a range of language and communication skills in AS/HFA. This review suggests that a comprehensive examination of not only semantic development, but also morphological and syntactic content, meaning, language structure, and social communication skills such as pragmatics, verbal problem solving, and abstract language skills (Manjiviona, 2003) may be required to extend the clinical understanding of language and communication skills associated with AS and HFA. No known study to date has applied the three theoretical approaches to investigating these language and communication skills in the disorders.

11 1.4.2

Language and Communication Skills in Adulthood

The autistic presentation in individuals is a life-long developmental disability (Hill & Frith, 2003), but as evident from the studies presented in section 1.3, most research has focused on autism in childhood. Language skills in ASD continue to change from adolescence to adulthood (Seltzer et al. 2003) and the maturational changes in language skill from childhood to adulthood are not, as yet, clearly understood in the disorder (Howlin, 1997). The current lack of awareness of the disorder in adulthood may have clinical implications. Howlin postulated that language and communication difficulties may be impediments to successful psychosocial outcomes in adults with ASD. If this is the case, defining the language and communication skills associated with the disorder is essential to determine if intervention and ongoing support is warranted in older adolescents and adults with a diagnosis of pervasive developmental disorder, whether it be AS, AD, HFA, or ASD. One of the three theoretical approaches to researching language in autism has been applied in studies investigating the language skills of adults with ASD. Howlin (2003) adopted the validation approach to examine the outcome of adults with autism, with and without a history of early language delay. She described marginal differences in receptive and expressive semantic abilities between those with AS and those with HFA. Further, Howlin described adults with either diagnosis as presenting with impaired linguistic skills relative to their chronological age. Rumsey and Hamburger (1988) likewise focused on adults with an autism diagnosis. Their study, however, constituted neither a validation nor a combined AS/HFA study. Rumsey and Hamburger focused on a diagnostically homogeneous group of adults with verbal and performance IQ above 80, and a childhood diagnosis of Infantile Autism, based on DSM-III criteria (APA, 1980). The authors noted some ill-defined linguistic deficits relative to typical peers. Rumsey and her co-researcher documented the heterogeneity of verbal skills within the homogeneous diagnostic group. They did not, however, extend their methodology to examine within-group differences.

12 It is clinically advantageous to extend the current awareness of the language and communication difficulties in adults with an autistic presentation, particularly if, as advocated by Howlin (1997), difficulties with language and communication negatively influence psychosocial wellbeing. Studies undertaken by Howlin (2003) and Rumsey and Hamburger (1988) focused on semantic skills only. As argued in section 1.4.1, however, studies are needed that extend beyond semantic investigations. Research investigating language and communication skills in adults with an autism diagnosis should encompass a wide range of communicative assessments. Such assessments should include a comprehensive examination of the skills suggested by Manjiviona (2003), as presented in section 1.4.1. Furthermore, if heterogeneous language skills within AS/HFA represent differing autistic presentations, intervention strategies may differ (Klin, 2000; Szatmari et al., 1986). The clinical utility of studies that have not investigated differences within the AS/HFA presentation may, inevitably, be limited. Hence, an examination of the diversity of language and communication skills in adults on the autism spectrum may be required to provide clinically applicable findings. 1.4.3

The Female Presentation

ASD can occur in both males and females (Baron-Cohen, 2002; Gillberg, 1989; Howlin & Asgharian, 1999; Kadesjo, Gillberg, & Hagberg, 1999; Wing, 1981). To date, there is a dearth of information on whether the language and social skills associated with females with a diagnosis of ASD are similar to the skills associated with males with the disorder. Many studies have included female participants (e.g., Channon, Charman, Heap, Crawford, & Rios, 2001; Ghaziuddin et al., 2000; Griswold, Barnhill, Myles, Hagiwara, & Simpson, 2002; Gunter, Ghaziuddin, & Ellis, 2002; Howlin, 2003; Rinehart, Bradshaw, Brereton, & Tonge, 2002; Rinehart, Bradshaw, Moss, Brereton, & Tonge, 2001; Starr, Szatmari, Bryson, & Zwaigenbaum, 2003; Szatmari et al., 2000), but no studies have undertaken within-group analyses to define the language and social skills in the female presentation relative to males with the disorder. While Frith (2003) and Kopp and Gillberg (1992) have posited that females with ASD may have better language and

13 social skills than males with the disorder, it is clinically imperative that the language and communication skills of females with ASD be defined to determine if females with the disorder require language intervention and support, and if so, is this support likely to be different from that needed by the male ASD population. 1.4.4

Influence of Cognition

In any discussion of language development and functioning, the dilemma of the interplay between cognition and language should, necessarily, be included. The exact nature of the interdependence between language and cognition is difficult to determine (Wetherby & Gaines, 1982). Mayes and Calhoun (2001) found no significant linguistic differences between their two groups of children with similar full scale, nonverbal, and performance IQ. Similarly, Howlin (2003) detected no major linguistic differences between her groups that were similar in nonverbal IQ. In contrast, Szatmari et al. (1995) described distinct language subgroups between their two groups of children whose nonverbal skills were significantly different. Due to incomplete data, Prior et al. (1998) were not able to examine the relationship between IQ and subgroup membership in their study. As yet, the influence of cognitive performance over language performance has not been clarified. It may be that cognitive skill is one marker differentiating individuals within the spectrum of autistic disorders. 1.5

Rationale for the Series of Studies

The hallmarks of clinically useful findings from research into AS and HFA are encapsulated in Gillberg’s (1998) management guidelines for AS and HFA, where the first guideline states “appropriate assessment [italics added] and correct diagnosis [is] essential for optimal understanding and service” [italics added] (p. 208). The current review suggests that, due to three methodological limitations, recent research findings may not have facilitated clinical advancement towards meeting Gillberg’s management guidelines. Firstly, the historical conceptualisation of AS and HFA being variations on a

14 theme has provided the theoretical basis for three differing approaches to research. All three approaches have examined aspects of language and/or communication skills associated with AS and HFA. As highlighted above, however, no approach to date has examined a comprehensive range of language and communication skills. To provide clinically relevant findings, standardised language and communication tests that investigate a wide range of language and communication skills are necessary. Secondly, research into language and communication skills in ASD has, to date, focused predominantly on children with the disorder. There are suggestions that the psychosocial wellbeing of adults with ASD may be reduced due to compromised language and communication difficulties (Howlin, 1997). Findings from previous studies have suggested linguistic difficulties may be experienced. It is imperative that comprehensive assessments be undertaken to accurately portray the language and communication skills associated with ASD in adulthood. Once skill strengths and deficits have been identified, specific intervention and support can be instigated if warranted. Thirdly, it has been postulated that females with ASD may present with better language and social skills than males with the diagnosis. This conjecture has not been tested, as no study has fully examined the language and communication skills in females with ASD. Gender differences in language and communication skills in ASD will have clinical relevance if the presentation between the sexes differs. 1.6

Aims of the Thesis

Primarily, the aim of the research is to extend the clinical understanding of the language and communication skills in children, adults, and females with a diagnosis of AS, HFA, or ASD (all diagnoses henceforth referred to as ASD unless otherwise stated) and with no documented intellectual impairment (henceforth referred to as average intelligence unless otherwise stated). The studies to be reported in this thesis will seek to provide findings that have clinical utility in assisting with determining intervention and support specifically for individuals of average intelligence with diagnoses of ASD. The framework for the current thesis is based on the three theoretical approaches to research

15 into the language and communication skills in AS and HFA identified in the literature review presented in this chapter. The clinical applicability of findings from the different theoretical approaches will be discussed. Specifically, this thesis aims to: •

Investigate the linguistic, complex abstract, and pragmatic language, and social problem solving abilities of a group of children of average intelligence with a diagnosis of ASD. These investigations are reported in chapters 2 - 5.

•

Investigate the linguistic, complex abstract, and pragmatic language, and social problem solving abilities of a group of adults of average intelligence with a diagnosis of ASD. These investigations are reported in chapters 6 - 8.

•

Determine whether language and social pragmatic skills differentiate the female and male adults with a diagnosis of ASD. This investigation is reported in chapter 9.

16

2

Linguistic Abilities in Children with Autism Spectrum Disorder 2

2.1

Introduction

In 1994, DSM-IV (APA) established a categorical diagnostic distinction between AS and AD (which includes HFA), with differential diagnosis related to reported developmental language history. Individuals with autistic symptomatology (i.e., abnormal communication, impaired social interaction and relationships, and rigid and limited imagination and play) (Rutter, 1978; Tantam, 1991; Wing, 1989) coupled with delayed onset or lack of language development are currently diagnosed as having AD. In contrast, individuals with autistic features combined with language onset prior to two years of age are currently diagnosed as having AS. Linguistic skills associated with the diagnostic categories of AS and HFA have been the focus of considerable research (e.g., Howlin, 2003; Mayes & Calhoun, 2001; Prior et al., 1998; Shields et al., 1996; Szatmari et al., 1995). Such research studies can be characterised by two of the three theoretical approaches outlined in chapter 1. One approach has been validation studies which have aimed to determine if AS and HFA are similar presentations, and whether a separate diagnostic entry of AS in the DSM-IV (APA, 1994) is warranted (e.g., Howlin; Mayes & Calhoun). The other approach has preempted conclusive findings from validation studies by viewing AS and HFA as similar disorders on an autism spectrum (e.g., Prior et al.). The provision of a diagnosis of AS or HFA allows access to intervention (Bishop,

2

“Linguistic abilities in children with Autism Spectrum Disorder,” by F. M. Lewis, B. E. Murdoch, and G.

C. Woodyatt, 2007, Research in Autism Spectrum Disorders, 1, 85-100, was adapted from this chapter.

17 2006). Beyond diagnosis, however, detailed knowledge of the linguistic competencies in ASD is necessary for the formulation of goals for intervention from educational support services. Clinically applicable research findings are, therefore, needed to enhance the clinical understanding of the autistic presentation by accurately defining the areas of strengths and difficulties. It may be, however, that previous findings of the linguistic skills in AS and HFA may not accurately reflect the difficulties experienced in ASD, as methodological limitations of the studies to date may have restricted the clinical application of findings. 2.1.1

Validation Studies

The inclusion of AS as a categorically distinct disorder in DSM-IV (APA, 1994) has resulted in ongoing debate regarding the external validity of the diagnosis of AS from the diagnosis of AD, particularly in terms of HFA (e.g., Frith, 2003; Klin et al., 2005; Klin & Volkmar, 2003; Leekam et al., 2000; Macintosh & Dissanayake, 2004; Mayes et al., 2001; Rutter, 1978). Numerous studies have examined linguistic outcomes in AS and HFA (Ghaziuddin et al., 2000; Howlin, 2003; Klin, Volkmar, Sparrow, Cicchetti, & Rourke, 1995; Mayes & Calhoun, 2001; Szatmari et al., 1995; Szatmari et al., 1990). Mayes and Calhoun, for instance, examined the outcomes of children with a history of early speech/language delay (autism) and children with a history of normal language onset and development (AS), but with similar full scale, nonverbal, and performance IQ. Data from clinical observations of the child, parent interviews, teacher reports, and previous evaluations were combined to complete the Checklist for Autism in Young Children (Mayes & Calhoun, 1999). No significant differences were noted on any of the expressive language measures between the two groups. Mayes and Calhoun’s (2001) results supported Szatmari et al.’s (1990) earlier findings. Szatmari and colleagues administered a neurocognitive test battery that included the Children’s Token Test (Di Simoni, 1977) to children diagnosed as AS or HFA, and found no differences between AS and HFA on the language measure. In contrast, Klin et al. (1995) described distinct linguistic profiles for the two disorders. Klin and colleagues assigned their child participants the diagnostic label of AS

18 or HFA using a modified version of ICD-10 (WHO, 1992). The authors determined that, despite a similar full scale IQ, neuropsychological profiles differentiated individuals with AS from HFA. Using measures of clinical observation rather than normed assessments, Klin et al. described qualitative differences between the two disorders on articulation, verbal output, vocabulary, and verbal memory. Szatmari et al. (1995), however, argued that quantitative differences alone differentiated AS and HFA on linguistic outcomes. The children investigated by Szatmari and colleagues who were diagnosed as HFA on the basis of their developmental language histories consistently performed two standard deviations below the mean on a range of language tests, while the group diagnosed AS on the basis of developmental language histories performed just below or within one standard deviation from the mean (see section 1.3.1 for the language tests administered). Unlike the methodologies employed by Mayes and Calhoun (2001) and Klin et al. (1995), however, Szatmari et al. did not match their participants on nonverbal ability. 2.1.2

Spectrum Approach

Szatmari et al.’s (1995) findings of quantitative linguistic differences accord with the alternative approach to defining linguistic skills in AS and/or HFA; that is, the conceptualisation of AS and HFA as disorders on an autism spectrum (Wing, 1981). Numerous studies have used the ASD terminology when defining experimental groups to examine the communication skills associated with the disorders (e.g., Paul et al., 2005; Prior et al., 1998; Seltzer et al., 2003; Warreyn et al., 2005). Few of these studies, however, have focused on linguistic skills. Prior et al. adopted a spectrum approach in their study by including diagnoses of AS, HFA and/or pervasive developmental disorder. Reflecting Wing’s (1989) proposal of an autism continuum, the authors examined withingroup differences. Based on performance on the British Picture Vocabulary Test (Dunn et al., 1982) or the Peabody Picture Vocabulary Test-Revised (Dunn & Dunn, 1981), Prior et al. described a number of subgroups differing on verbal skills, but due to incomplete data, were not able to examine the relationship between IQ and subgroup membership in their study.

19 2.2

Rationale for the Present Study

Research findings gain relevance when they are applied in clinical practice, thus providing an evidence-base for treatment and support (Rutter, 2005). The clinical applicability of findings from validation studies and the spectrum approach taken by Prior et al. (1998) may be restricted, however, due to the theoretical perspective adopted in both approaches. Linguistic testing for the validation studies and the spectrum study was restricted to basic linguistic skills, such as expressive and receptive vocabulary and syntax. Theoretically, this position aligns with DSM-IV’s (APA, 1994) focus on early semantic and syntactic development, which has diagnostic relevance only. Knowledge of an individual’s diagnosis, whether it be AS or HFA, however, offers little information for planning effective support programmes (Griswold et al., 2002). The subsequent development of communicative competence, which encompasses broader aspects of language development such as morphological and syntactic content, meaning, and structure (Semel et al., 2003) does not have diagnostic relevance, yet these aspects are highly relevant in terms of the provision of clinical support. 2.3

Aims and Hypotheses of the Present Study

The present study aimed to assess the linguistic skills in children with a diagnosis of ASD in order to provide research findings that have clinical application in the determination of the type and level of language intervention and support needed following diagnosis. To extend previous findings, linguistic skills beyond those central to the diagnostic process were included in the linguistic assessment. The first aim was to examine whether current linguistic performance differentiated AS from HFA. Based on the methodology applied in the validation studies, the ASD children were reclassified as AS or HFA according to DSM-IV’s (APA 1994) language criterion. Between-group differences were then examined. The second aim was to administer a comprehensive assessment of semantic, morphological, and syntactic content, meaning, and language structure to children with a diagnosis of AS, HFA, or ASD. Based on the methodology used in the combined studies discussed in chapter 1, no delineation of the diagnostic groups was attempted. Group performance was compared to a typically developing control group. Finally,

20 incorporating aspects of the methodology applied by Prior et al. (1998) in their spectrum study, hierarchical cluster analysis of the combined ASD participants’ performance on the language measures was undertaken to investigate subgroups within ASD. No hypothesis was made regarding a differentiation on linguistic measures of the ASD children when reclassified as AS and HFA as previous conclusions (e.g., Ghaziuddin et al., 2000; Howlin, 2003; Klin et al., 1995; Mayes & Calhoun, 2001; Szatmari et al., 1995; Szatmari et al., 1990) have been based on restricted semantic/grammatical assessments only. Based on previous findings of linguistic deficits in both AS and HFA, it was hypothesised that the non-delineated ASD group would present with significantly different linguistic skills when compared to a control group comprised of normally developing peers. Finally, as previous investigations of withingroup analyses have identified subgroups reflecting the heterogeneous presentation within ASD, it was hypothesised that subgroups, based on current language status, would be evident. 2.4

Method

2.4.1

Participants

A total of 20 children/adolescents with average intelligence [as evidenced by no previously documented intellectual impairment] (16 male, 4 female; M age: 11 years 6 months; SD: 2 years 2 months; age range: 9 years 0 months – 17 years 1 month; M years of schooling: 6 years 1 month; education years range: 3 years – 11 years) with a diagnosis of AS, HFA, and/or ASD were recruited through the Asperger Syndrome Support Network of Queensland, Autism Queensland, and newspaper articles. Recruitment articles specified that the children and adolescents must have a formal diagnosis of AS, HFA, or ASD. Table 2.1 (following) displays the demographic details of the participants. In addition to a diagnosis of AS/HFA/ASD, inclusion criteria for the study were: no formal diagnosis of intellectual impairment, aged nine years or older, English as a first language, no neurological disease or trauma, no other co-morbid condition, no history of

21 drug and/or alcohol abuse, and normal vision and hearing. Due to financial constraints, no formal IQ testing was undertaken for the study. Documentation of diagnosis was requested before the commencement of testing. Diagnoses made by psychiatrists, paediatricians, and psychologists were accepted for inclusion into the study. A range of diagnostic criteria have been proposed for AS (e.g., APA, 1994; Gillberg & Gillberg, 1989; Szatmari, Bremner, & Nagy, 1989; WHO, 1992) and the criteria used by the various professionals for diagnosis of participants were not known. It is acknowledged that some of the diagnoses may not adequately reflect DSMIV criteria in regards to developmental language histories. Refer to Appendixes A and B for DSM-IV criteria for AD and AS respectively. A total of 18 children/adolescents (14 males, 4 females; M age: 11 years 5 months; SD: 1 year 8 months; age range: 9 years 1 month – 14 years 8 months; M years of schooling: 6 years 2 months; education years range: 4 years – 9 years) with nonsignificant developmental histories were recruited as control participants. A student t test indicated there was no significant difference in age between the ASD and control group. Recruitment of control participants was through newspaper articles in local papers, local sporting clubs, and schools in the south-east region of the state of Queensland, Australia. To be included in the control group, participants were required to be aged nine years or older, have English as a first language, no history of neurological trauma or disease, no history of drug and/or alcohol abuse, and normal hearing and vision. To avoid the issues of co-morbidity with autism and/or language difficulties, no family history of autism, and no developmental or family history of language delays or difficulties were also required for control participant inclusion.

22 Table 2.1 ID

Sex

ASD child participant demographics Age in

(M: 16; years, Fe: 4)

Dev lge

Given diag

Age at

Reclassified

hist

(Source)

diagnosis diagnosis

Early intervent

(in years) using DSM-

months

IV language criterion 1

M

9, 4

Normal

AS (paed)

8

AS

Nil

2

M

11, 11

Normal

ASD (paed)

8

AS

Nil

3

M

14, 0

Normal

AS (psych)

12

AS

Nil

4

M

11, 0

Normal

ASD (paed)

7

AS

Nil

5

M

12, 1

Normal

AS (paed)

11

AS

Lge

6

M

12, 6

Normal

AS (psychiat)

8

AS

Nil

7

Fe

13, 6

Delayed

AS (psychiat)

11

HFA

Lge

8

M

17, 1

Normal

AS (paed)

8

AS

Lge

9

M

15, 1

Normal

AS (paed)

12

AS

Sp; Lge

10

Fe

9, 7

Delayed

AS (paed)

8

HFA

Sp; Lge

11

M

13, 7

Delayed

AS (paed)

12

HFA

Lge

12

M

9, 5

Unsure

AS (paed)

8

Unsure

Lge

13

M

11, 6

Delayed

AS (psychiat)

7

HFA

Sp; Lge

14

M

10, 5

Delayed

AS/ASD (paed) 6

HFA

Beh; Lge

15

M

10, 10

Normal

AS (psych)

AS

Lge

16

Fe

13, 3

Delayed

AS/ASD (paed) 6

HFA

Lge

17

M

9, 8

Delayed

ASD (paed)

3

HFA

Lge

18

M

9, 0

Delayed

ASD (paed)

7

HFA

Lge

19

Fe

9, 0

Unsure

ASD (paed)

5

Unsure

Lge

20

M

10, 5

Normal

ASD (paed)

5

AS

Nil

6

Table continues

23 Table 2.1 continued Note. ASD = Autism spectrum disorder. AS = Asperger syndrome. HFA = High functioning autism. Dev lge hist = Developmental language history. Normal = Language onset prior to age two years. Delayed = Language onset after two years of age. Unsure = Informant unable to provide details. Given diag = Independently determined diagnosis. Source = Source of diagnosis. M = Male. Fe = Female. Psych = Psychologist. Paed = Paediatrician. Psychiat = Psychiatrist. Reclassified diagnosis = Based on developmental language histories provided, participants were reclassified as AS or HFA using DSM-IV’s language criterion. Early intervent = Focus of early intervention. Lge = Language intervention. Sp = Speech/articulation intervention. Beh = Behavioural intervention.

24 2.4.2 Measures The Clinical Evaluation of Language Fundamentals-Fourth Edition (CELF-4) (Semel et al., 2003) and the Test of Nonverbal Intelligence-Second Edition (Form A) (TONI-2) (Brown, Sherbenou, & Johnsen, 1990) were administered to all participants. The CELF-4 (Semel et al., 2003) is a standardised test for the paediatric population, and assesses the basic foundations of mature language use. The test assesses language competence through an examination of semantic content, morphological and syntactic content, meaning, and structure. The CELF-4 subtests administered included: Concepts and Following Directions, Recalling Sentences, Formulated Sentences, Word Classes, Expressive Vocabulary, Word Definition, Understanding Spoken Paragraphs, Sentence Assembly, and Semantic Relationships. The subtests administered depended on the child’s age. The CELF-4 provides Index Scores that are calculated by using the scores of subtests that measure similar language features (Semel et al.). The five Index Scores computed for the present study were Core Language, Receptive Language, Expressive Language, Language Content, and Language Memory. Statistical properties of a test determine the confidence placed in the accuracy and precision of the obtained scores (Semel et al., 2003). The psychometric properties provided in the CELF-4 manual are adequate. The average corrected stability coefficients for all ages for the various subtests range from .70-.90. Average reliability coefficients of the CELF-4 subtests range from .71-.92 and the average reliability coefficients of the CELF-4 Index Scores range from .87-.95. The TONI-2 is a language-free measure of cognitive ability (Brown et al., 1990). It was administered to determine if differences in verbal results were due to nonverbal cognitive abilities rather than linguistic factors only. The TONI-2 technical manual includes psychometric properties of the test. The average internal consistency reliability for Form A using the coefficient Alpha formula is satisfactory (.95). Test validity, using correlations between TONI-2 scores and a range of measures of aptitude and general intelligence exceed statistically significant coefficients (.35) for the majority of measures. The TONI-2 was administered as an indicator of nonverbal cognitive skill, not as an overall marker of intelligence.

25 2.4.3 Procedure All parents or caregivers of the participants completed a questionnaire prior to the commencement of testing. As well as providing identifying data, the ASD questionnaire requested a history of language development. Questions were directed towards the timing of the onset of single words and the development of simple communicative phrases reflecting DSM-IV’s (APA, 1994) language criterion. The questions were worded in a way to avoid parents/caregivers making decisions regarding the appropriateness of the timing of language onset. Where parents relied on retrospective recall of language milestones, prompts were given to link language onset with other milestones such as birthdays. Appendix C displays aspects of the questionnaire relating to developmental language history. Judgments regarding language onset were made by an experienced speech pathologist (the author). Specifying the timing of the onset of simple communicative phrases proved difficult for many parents/caregivers and the data set was deemed too unreliable to include in the analyses. Judgment for classifying language onset was, therefore, based on DSM-IV’s language criterion (APA, 1994) for single words (i.e., onset of single words prior to age 2 years = normal onset = AS; onset of single words after age 2 years = delayed onset = HFA). Inter-rater reliability of 100 % agreement on decisions was provided by a second experienced speech pathologist. Where language history could not be provided, this was documented as unsure. Developmental language history was not available for two children. For one child, the parents were not able to recall developmental milestones, and for the other, developmental history was not available as the child did not live with his family of origin. Children with a diagnosis of ASD were reclassified as AS or HFA based on their reported developmental language history. Using DSM- IV’s (APA, 1994) language criterion (see previous paragraph), the reclassification of the children resulted in three groups: AS (onset of first words prior to two years of age): n = 10, age range: 9 years 4 months – 17 years 1 month, HFA (onset after two years of age): n = 8, age range: 9 years 0 months – 13 years 7 months, and those with unsure developmental language history (n = 2, age range: 9 years 0 months – 9 years 5 months). Refer to Table 2.1 for the reclassified diagnoses. The Unsure group data were excluded from the reclassification statistical analyses. A Mann-

26 Whitney U test indicated there was no significant difference in age between the AS and HFA group (Z = -1.022; p = .307). The Behavioural and Social Sciences Ethical Review Committee at the University of Queensland, Australia gave ethical approval to the study involving the children. Signed informed consent was required from child participants and their parents or caregivers prior to acceptance into the study. Testing for this and subsequent aspects of the research involving the children was undertaken at the University of Queensland, at the participant’s school, or in the home of the participant. Each assessment was administered in a distraction-free environment in a standardised manner according to the instruction manuals. No time limits were placed on participant responses, unless specified by the testing manual. 2.5

Results

2.5.1 Performance on the CELF-4 by the Children with ASD Reclassified as AS or HFA Based on Reported Developmental Language History Means and standard deviations for the AS and HFA groups are shown in Table 2.2. For determination of statistical significance, the level of p ≤ .05 was utilised. Comparative analyses of the children were undertaken using Mann-Whitney U tests. No significant differences were detected between the two groups on age (p ≥ .05), nonverbal intelligence (p ≥ .05), Core Language (p ≥ .05), Receptive Language (p ≥ .05), Expressive Language (p ≥ .05), Language Content (p ≥ .05), or Language Memory (p ≥ .05).

27 Table 2.2 Descriptive statistics of the performance on the CELF-4 by children with ASD reclassified as AS or HFA according to developmental language history

Variable

Language onset

Language onset

Mann-Whitney U

prior to two

after two years

Test

years of age

of age

(AS)

(HFA)

(n = 10)

(n = 8)

M (SD)

M (SD)

Z

p (Asymp. Sig. 2tailed)

Age (yrs, mths)

12, 5 (2, 4)

11, 4 (1, 11)

-1.022

.307

TONI-2

103 (18)

99 (20)

-0.867

.386

Core Lge

91 (20)

77 (22)

-1.200

.230

Rec Lge

93 (19)

82 (20)

-1.112

.266

Exp Lge

92 (21)

80 (20)

-1.068

.285

Lge Cont

91 (15)

77 (18)

-1.605

.108

Lge Mem

88 (20)

79 (24)

-1.156

.248

CELF-4

Note. ASD = Autism spectrum disorder. AS = Asperger syndrome. HFA = High functioning autism. TONI-2 = Test of Nonverbal Intelligence-2nd Edition. CELF-4 = Clinical Evaluation of Language Fundamentals-4th Edition. Core Lge = Core Language Index from CELF-4. Rec Lge = Receptive Language Index. Exp Lge = Expressive Language Index. Lge Cont = Language Content Index. Lge Mem = Language Memory Index. Yrs, mths = Years, months.

28 2.5.2

Performance by all Child Participants on the CELF-4

Group means and standard deviations for all ASD and control participants are shown in Table 2.3. There was no significant difference between the two groups on age (p ≥ .05). A significant difference between the two groups was detected on nonverbal intelligence (p ≤ .05). For each measure, a Levene’s test was carried out, and where necessary, appropriate corrections for non-homogeneity of variance were undertaken. Student t tests indicated that the children with ASD performed significantly less well than the normally developing children on Core Language (p ≤ .001), Receptive Language (p ≤ .05), Expressive Language (p ≤ .001), Language Content (p ≤ .001), and Language Memory (p ≤ .01). Table 2.3

Variable

Descriptive statistics of the child participants on the CELF-4 ASD

Control

(N = 20)

(N = 18)

M

t

Sig. (2tailed)

SD

M

SD

Age (yrs, mths) 11, 6

2, 2

11, 5

1, 8

0.223

.825

TONI-2

100

18

113

16

-2.259

.03 *

Core Lge

84

20

106

12

-3.800

.001 ***

Rec Lge

88

19

98

10

-2.169

.037 *

Exp Lge

86

20

107

12

-3.943

.001 ***

Lge Cont

84

17

102

14

-3.729

.001 ***

Lge Mem

84

21

102

12

-3.428

.002 **

CELF-4

Note. ASD = Autism spectrum disorder. TONI-2 = Test of Nonverbal Intelligence-2nd Edition. CELF-4 = Clinical Evaluation of Language Fundamentals-4th Edition. Core Lge = Core Language Index from CELF-4. Rec Lge = Receptive Language Index. Exp Lge = Expressive Language Index. Lge Cont = Language Content Index. Lge Mem = Language Memory Index. Yrs, mths = Years, months. * = p ≤ .05. ** = p ≤ .01. *** = p ≤ .001.

29 Descriptively, the standard deviations for the ASD group are much larger than the control group on Core Language, Receptive Language, Expressive Language, and Language Memory (refer to Table 2.3), suggesting a greater dispersion of scores for the ASD group. Through the use of box plots, Figures 2.1 - 2.5 display the dispersion of the five CELF-4 Index Scores for the two data sets. Box plots are graphs that reflect the central tendency and the variability of a distribution of a variable. The middle 50 % of the data sets is displayed in the shaded part of the box plots, with the lower hinge representing the 25th percentile and the upper hinge representing the 75th percentile. The horizontal line within the box plot is the median. The short horizontal lines outside the box represent the largest values that would not be regarded as outliers. Values beyond these lines are regarded as outliers. An outlier is marked by an open circle and a far outlier is marked by an asterix (Lane, n.d.). The length of the box plot will be greater for samples with more dispersion of scores (Weinbach & Grinnell, 1998). Performance by one ASD child on Receptive Language was identified as an outlier. The dispersal of scores for the ASD group is greater than the control group on the Core Language, Receptive Language, Expressive Language, Language Content, and Language Memory Index Scores.

30

140

120

100

Core

80

Lge 60

40

N=

20

18

ASD

Control GROUP

Figure 2.1.

Box plots displaying the dispersal of scores from the child ASD

and control groups’ performance on the Core Language Index from the Clinical Evaluation of Language Fundamentals-4th Edition.

31

140

120

100

Rec

80

Lge 60

10

40

N=

20

18

ASD

Control GROUP

Figure 2.2.


and control groups’ performance on the Receptive Language Index from the Clinical Evaluation of Language Fundamentals-4th Edition.

32

140

120

Exp Lge

100

80

60

40

N=

20

18

ASD

Control GROUP

Figure 2.3.


and control groups’ performance on the Expressive Language Index from the Clinical Evaluation of Language Fundamentals-4th Edition.

33

120

Lge Cont

100

80

60

40

N=

20

18

ASD

Control GROUP

Figure 2.4.


and control groups’ performance on the Language Content Index from the Clinical Evaluation of Language Fundamentals-4th Edition.

34

140

120

Lge Mem

100

80

60

40

N=

20

18

ASD

Control GROUP

Figure 2.5.


and control groups’ performance on the Language Memory Index from the Clinical Evaluation of Language Fundamentals-4th Edition.

35

2.5.3

Clusters of Children with ASD Based on Performance on the CELF-4

To investigate potential language subgroupings within the ASD group, performance scores from the five Index Scores provided by the CELF-4 were entered for an agglomerative hierarchical cluster analysis. In agglomerative hierarchical clustering, individual cases are initially considered a cluster. The two cases with the lowest distance (or highest similarity) are then combined into a cluster. The case with the lowest distance to either of the first two is considered next. If that third case is closer to a fourth case than it is to either of the first two, the third and fourth cases become the second two-case cluster. If the third case is not closer to the fourth case, it is added to the first cluster. The process is repeated for each case, adding to existing clusters, creating new clusters, or combining clusters to get the final number of clusters. The rescaled distance refers to similarity in measures, and cases with a similar rescaled distance cluster early. Greater distances at which clusters combine denote greater heterogeneity within the clusters. When a horizontal line joins two vertical portions, it indicates fusion between two clusters of cases (NCSU, n.d.; StatSoft, 1984-2004). Average linkage clustering (Sokal & Michener, 1958) was used for the analysis. Average linkage clustering is a relatively robust method of analysis (Everitt, Landau, & Leese, 2001) which computes an average of the similarity of a case under consideration in regards to all cases in the existing cluster. If there is a level of similarity using this average value, the case is joined to the existing cluster (Aldenderfer & Blashfield, 1984). The dendogram is a pictorial representation of the complete clustering procedure. There are a number of approaches to determining the number of groups, but Baxter (1994) has suggested informal and subjective criteria remain the most common approach. The hierarchical clustering of the ASD children is shown in Figure 2.6.

36

Figure 2.6.

Agglomerative hierarchical cluster analysis of the 20 ASD children

based on performance on the five Index Scores from the Clinical Evaluation of Language Fundamentals-4th Edition.

37 The partitioning of the dendogram into clusters was based on informal and subjective analysis. Four subgroups can be identified from the cluster analysis. The resultant subgroups are descriptive only. Table 2.4 displays the membership of the subgroups. Each participant’s score is defined as within a number of standard deviations from the control group’s mean. Performance scores within each subgroup support the partitioning decision.

38

Table 2.4 Subgroups of ASD children as determined by the CELF-4 Index Scores showing standard deviations from the control group means ID

Given diag

Lge hist

Core Lge

Rec Lge

Exp Lge

Lge Cont

Lge Mem

M age: 11, 1 (yrs, mths)

16

AS/ASD

Delayed

-3.0

-2.5

-3.0

-2.5

-3.0

(SD: 2, 4 yrs, mths)

17

ASD

Delayed

-3.5

-2.5

-3.5

-2.5

-3.5

AS

Normal

-2.5

-2.5

-3.5

-2.5

-2.5

ASD

Unsure

-2.5

-1.5

-3.5

-2.0

-2.5

1

AS

Normal

-2.5

-2.0

-2.0

-2.5

-2.0

12

AS

Unsure

-2.0

-2.0

-1.5

-2.0

-1.0

5

AS

Normal

-2.5

-2.0

-2.0

-2.0

-4.5

Subgroup 2 (n = 6)

11

AS

Delayed

-1.5

-1.5

-1.5

-1.5

-0.5

M age: 12, 2 (SD: 8, 0)

14

AS/ASD

Delayed

-1.0

-0.5

-1.0

-1.5

-1.5

6

AS

Normal

-1.5

-1.5

-1.0

-0.5

-1.5

15

AS

Normal

-2.0

-1.5

-1.5

-0.5

-2.0

18

ASD

Delayed

-2.0

-1.5

-2.0

-1.0

-2.0

AS

Normal

-2.0

0.0

-2.5

-0.5

-1.5

Subgroup 1 (n = 7)

M TONI-2: 92 (SD: 11)

9 19

M TONI-2: 107 (SD: 15)

8

Table continues

Table 2.4 continued ID

Given diag

Lge hist

Core Lge

Rec Lge

Exp Lge

Lge Cont

7

AS

Delayed

-4.5

-4.0

-4.0

-3.5

-4.0

M age: 11, 1 (SD: 2, 0)

10

AS

Delayed

-5.0

-5.0

-4.5

-4.0

-4.5

M TONI-2: 81 (SD: 11)

20

ASD

Normal

-4.0

-3.5

-4.0

-2.0

-3.5

4

AS

Normal

+0.5

+2.0

+1.0

+0.5

+1.0

13

AS

Delayed

+0.5

+1.5

+0.5

0.0

+1.0

2

ASD

Normal

+0.5

+1.5

+1.5

+1.0

+1.5

3

AS

Normal

+1.0

+2.5

+1.5

+0.5

+2.0

Subgroup 3 (n = 3)

Subgroup 4 (n = 4) M age: 12, 1 (SD: 1, 3) M TONI-2: 120 (SD: 12)

Lge Mem

Note. ASD = Autism spectrum disorder. ID = Identification number. Given diag = Independently determined diagnosis. AS = Asperger syndrome. Lge hist = Developmental language history. Normal = Language onset prior to age two years. Delayed = Language onset after two years of age. Unsure = Developmental language history not provided. CELF-4 = Clinical Evaluation of Language Fundamentals-4th Edition. Core Lge = Core Language Index from CELF-4. Rec Lge = Receptive Language Index. Exp Lge = Expressive Language Index. Lge Cont = Language Content Index. Lge Mem = Language Memory Index. TONI-2 = Test of Nonverbal Intelligence-2nd Edition. Yrs, mths = Years, months. Scores are entered as within standard deviations above or below the control means. + = above the control mean. - = below the control mean.

39

40 Subgroup 1 was characterised by mild to moderate difficulties on Receptive Language and Language Content, and moderate-severe difficulties on Core Language, Expressive Language, and Language Memory. Subgroup 2 presented with generally mild difficulties on Receptive Language and Language Content, and mild to moderate difficulties on Core Language, Expressive Language, and Language Memory. Subgroup 3 presented with severe deficits across all five CELF-4 Index Scores, and Subgroup 4’s performance was average to above average on all five CELF-4 Index Scores. Non-parametric tests (Kruskal-Wallis) established there was no significant difference between the four ASD subgroups on age (χ2 = 1.426; p = .700), but a significant difference on the TONI-2 (χ2 = 10.317; p = .016). Mann-Whitney U tests determined that there were significant differences in performance on the TONI-2 between subgroups 1 and 4 (Z = 1.000, p = .014) and 3 and 4 (Z = 0.000, p = .034). There were no significant differences between subgroups 1 and 2 (Z = 9.500, p = .099), 1 and 3 (Z = 4.500, p = .169), 2 and 3 (Z = 2.000, p = .071), or 2 and 4 (Z = 5.000, p = .136). 2.6

Discussion

The three theoretical frameworks developed from the historical link between AS and AD (presented in section 1.3) were used in the present study to investigate the linguistic capabilities associated with ASD in childhood. The first approach investigated categorical distinctions between AS and HFA, where reclassification of the ASD participants as AS or HFA using DSM-IV’s (APA, 1994) language criterion was undertaken. Reclassification failed to detect significant differences between the resultant groups on linguistic measures. In the second approach, children with a diagnosis of AS, HFA, or ASD were combined into a non-delineated ASD group. The combined group of children presented with linguistic deficits relative to typically developing peers, but this approach failed to examine the heterogeneity of linguistic capabilities noted within the ASD sample. The final approach, where within-group analysis was undertaken, yielded a more accurate presentation of the range of linguistic skills associated with ASD.

41 2.6.1

Linguistic Skills in Children with ASD Reclassified as AS or HFA According to Reported Developmental Language History

The regrouping of individuals with ASD using the reported timing of language onset (e.g., Ghaziuddin et al., 2000; Howlin, 2003; Mayes & Calhoun, 2001; Szatmari et al., 1995; Szatmari et al., 1990) was undertaken to determine if subsequent linguistic skills differed significantly between the groups designated AS and HFA based on DSM-IV’s (APA, 1994) language criterion (i.e., onset prior to age 2 years = AS, onset after age 2 years = HFA). In the present study, linguistic performance failed to differentiate the resultant AS and HFA groups. This finding is consistent with previous studies (e.g., Howlin; Mayes & Calhoun), where the timing of the onset of language was shown to be irrelevant in subsequent linguistic outcomes. Findings from validation studies, however, should be viewed cautiously as concerns have been raised regarding the reliability and validity of diagnostic decisions based on developmental history (Prior et al., 1998), which is the basic premise of such studies. The application of DSM-IV’s language criterion (APA, 1994) relies on the retrospective recall of language milestones. This diagnostic method is open to question, as it cannot be assumed that recall of developmental milestones is always accurate (Tager-Flusberg, 2003; Woodbury-Smith, Klin, & Volkmar, 2005). This is particularly applicable to AS. While some children are identified as young as three years of age (Semrud-Clikeman & Hynd, 1990), the average age at diagnosis is eight years (Attwood, 1998), with some individuals reaching adulthood before a diagnosis of AS is given (Frith, 2003). In such cases, developmental history may be unreliable or unavailable. Further, one group of researchers (Lord et al., 2000) has argued that developmental history and parent report should be considered in tandem with scores from a standardised observation of current behaviours such as the Autism Diagnostic Observation Schedule-Generic (ADOS-G) (Lord et al.). Use of scores from an instrument such as ADOS-G provides information on current social communication behaviour which assists the diagnostic process, but does not adequately address the ongoing diagnostic issue of the DSM-IV (APA, 1994) reliance on developmental history.

42 An additional concern regarding DSM-IV (APA, 1994) is that the interpretation of reported developmental language history may lead to conflicting diagnoses (WoodburySmith et al., 2005). For instance, DSM-IV’s language criterion for AS, that of the use of single words by two years of age and the use of communicative phrases by three years of age, would be viewed as delayed onset and development if using normative data from typically developing children (e.g., Tager-Flusberg, 2003). For some practitioners, such a presentation would be classified, not as AS, but as AD, due to delayed onset of speech and language. The reliability of studies that have designated participants into AS and HFA based on language onset as the participant inclusion criterion must, therefore, be questioned. Given Mayes and Calhoun (2001), for instance, classified their participants using developmental histories, their results may, therefore, be unreliable. There are also concerns regarding the interpretation of findings from validation studies, as the DSM-IV (APA, 1994) distinction between AS and HFA may not be applied in diagnostic and clinical practice (Klin et al., 2005). Despite the categorical distinction established in DSM-IV, Klin and Volkmar (2003) cautioned that the terms AS and HFA continue to be used synonymously. The resultant overlap in terms has important research implications, as unreliable diagnoses may lead to contaminated research findings. These problems were evident in the current study. The developmental histories of two children were unavailable. Further, there was evidence in the present study of a lack of adherence to DSM-IV’s (APA, 1994) language criterion, or possibly inconsistent interpretation of reported milestones, resulting in unreliable independently given diagnoses. For example, eight children with independently given diagnoses of AS were reclassified in the study as HFA when strict adherence to DSM-IV’s language criterion was applied. The findings of the current study suggest that diagnosticians such as psychiatrists, paediatricians, and psychologists may be disregarding developmental language history when making diagnostic decisions regarding autistic classification. To offer research utility, accuracy and clarity are needed in defining research populations to allow for comparisons across studies and for the replication of studies (Bishop, 2006; Klin & Volkmar, 2003). To offer clinical utility, accurate and consistent

43 diagnostic decisions are needed, as access to services may depend upon the given diagnosis (Bishop). Likewise, qualitatively different impairments may need different intervention strategies (Klin, 2000; Szatmari et al., 1986). Consequently, the utility of research findings from validation studies to extend the understanding of the linguistic outcomes in AS and HFA may be limited in clinical practice for practitioners, such as speech-language pathologists, providing intervention in school-based settings. The debate regarding differences between AS and HFA may remain unresolved while the reliance upon the timing of the onset of language to differentially distinguish between a diagnosis of AS and a diagnosis of HFA remains. 2.6.2

Linguistic Skills in Children with ASD

Previous studies have combined AS and HFA participants with the assumption there is no difference between the two autistic presentations (see e.g., Booth et al., 2003; Emerich et al., 2003; Losh & Capps, 2003; Ozonoff & Miller, 1996; Prior & Hoffmann, 1990; Rumsey, 1985; Rumsey & Hamburger, 1988; Shields et al., 1996). Adopting this theoretical approach, the present study found that, compared to typically developing peers, the children with ASD had difficulties on a wide range of linguistic parameters. The group of children with ASD presented with deficits in core language, receptive language, expressive language, language content, and language memory. The identification of linguistic deficits beyond merely semantic knowledge, such as deficits in language memory and content, suggests that children with ASD may require a range of supports in educational settings. Consistent with previous researchers (e.g., Emerich et al., 2003; Rumsey, 1985; Rumsey & Hamburger, 1988), heterogeneity within the ASD group was noted in the present study. Tager-Flusberg (2004), however, has maintained that variance within combined AS/HFA studies may hide potentially significant group differences on aspects of language performance. The findings of the combined ASD approach do not clarify the language and communicative status of individuals with diagnoses of AS or HFA, and therefore, may have limited clinical application in terms of guiding the development of intervention and support services. An examination of within-group differences may be

44 needed before combined ASD studies offer clinically relevant findings. 2.6.3

Heterogeneity of Linguistic Skills in Children with ASD

A number of studies have pre-empted conclusions from validation studies, and have viewed AS and HFA as disorders on an autism spectrum (e.g., Mandell et al., 2005; Prior et al., 1998). Importantly, this shift avoids the limitations associated with reliance upon developmental language history to differentiate AS and HFA, as presented in section 2.6.1. Few studies, however, have examined within-group differences, despite the acknowledged heterogeneity of the ASD population (e.g., Emerich et al., 2003; Griswold et al., 2002) and the conceptualisation of ASD as a continuum of autistic symptomatology (Wing, 1989). An examination of within-group differences in the current study revealed linguistic competence in the children with ASD ranged from above-average skills through to severe deficits on all language Index Scores. It is possible that the range of skills demonstrated in the present study represent variations as observed in the normal population. Normal variation, however, cannot account for the severe deficits observed in a number of the subgroups. Nonverbal cognition was a significant factor in the subgroup membership, suggesting the possibility that nonverbal cognitive skill may influence linguistic competence. The exact nature of the relationship between cognition and language is yet to be clarified (see e.g., Howlin, 2003; Klin et al., 1995; Mayes & Calhoun, 2001; Szatmari et al., 1995). It may be that cognitive skill, rather than the timing of the onset of language, is a marker that differentiates individuals on the autism spectrum. The description of subgroups based on performance on a comprehensive linguistic assessment extends the findings by Prior et al. (1998) who described subgroups within ASD based on performance on expressive semantics. The utility of examining withingroup differences in the ASD group is highlighted by the identification of groups of children with varying clinical needs. The range of competencies identified in the majority of the participants suggests that language difficulties may interfere with success in home and school life for children with ASD. The Expressive and Receptive Language Index

45 Scores from the CELF-4 (Semel et al., 2003) reflect basic language proficiency. Expressive language difficulties may reduce the effectiveness in oral and written expression through poor use of grammar and restricted word use. Reduced receptive language skill, as noted in some individuals, may result in difficulties following directions, understanding verbal concepts, and understanding spoken and written paragraphs. Deficient language memory skills may result in poor recall of complex spoken directions. Language content encompasses more complex skills such as category formulation, comprehending relationships between words, and interpreting inferential and factual information. Seventy percent of the children with ASD investigated in the current study had previously received speech pathology intervention for language difficulties. None were undergoing speech pathology intervention at the time of the present assessments. The current findings suggest that the majority of the ASD participants involved in the research have ongoing language intervention needs. One subgroup of children (subgroup 4) presented with above-average language skills. Within this subgroup, one child had previously received speech pathology intervention. His placement in the subgroup with above-average language skills suggests successful intervention in aspects of language development. The children in this subgroup would have no need of language support. Children in the three other clinical subgroups, however, would require considerable ongoing language support and intervention. The study’s findings suggest that monitoring language competence following diagnosis and the formulation of specific individual intervention plans would be required to optimise language skills to reduce the negative academic, social, and emotional outcomes for the majority of the children with ASD involved in the study. A limitation to this study, however, is that the analysis of errors that can be undertaken with the CELF-4 to highlight error patterns which can then be used to develop intervention plans (Semel et al., 2003), was not undertaken. 2.7

Conclusion

The three approaches adopted in the current study, which combine methodological

46 aspects from previous studies, offer findings relevant to both research and clinical practice. The finding that children with ASD could not be differentiated on linguistic performance when reclassified as AS or HFA using DSM-IV’s (APA, 1994) language criterion supports previous research results (e.g., Mayes & Calhoun, 2001). The methodological constraints inherent in using developmental language history as the defining categorical distinction between the two disorders, however, may limit the diagnostic and clinical potential of this finding. Relative to normally developing peers, the children grouped as an undifferentiated ASD group presented with a broad range of language deficits, encompassing overall language skills, receptive and expressive language, language content, and language memory. Due to considerable variation in performance scores within the ASD group, viewing AS and HFA as a single clinical group may provide little information that can be applied clinically in the formulation of intervention programmes. The examination of subgroups within ASD based on severity of linguistic difficulties has important therapeutic, but not diagnostic, implications. One subgroup presented with above-average skills across all language Index Scores. All five Index Scores were found to be impaired to varying degrees in the other three ASD subgroups. Linguistic competence for the children with ASD involved in the present study ranged from above-average performance through to severe difficulties. The range of linguistic skills described in the ASD group involved in the study suggests it is imperative that those identified as having ASD undergo a comprehensive linguistic assessment to allow for specific assistance tailored to their individual needs. The results of the current study suggest that a dimensional view of ASD, rather than the categorical approach as stipulated by DSM-IV (APA, 1994), may provide clinically relevant research findings which can be utilised to form the basis for management planning. A dimensional view of ASD necessitates a comprehensive language assessment for each diagnosed individual, which then facilitates individualised planning of language support and intervention. Studies that further examine the relationship between linguistic competence and cognition, however, would be of interest,

47 as cognitive skill may be more reliable than developmental history as a marker differentiating individuals within the spectrum of autistic disorders. Linguistic difficulties may contribute to many of the behavioural, social, and academic difficulties for children with a diagnosis of ASD (Lord, 2000). The identification of specific areas of difficulty, such as receptive language, expressive language, language content, and language memory can assist with the formulation of intervention goals. There are, however, other language skills that may likewise contribute to the behavioural, social, and academic outcomes for children with the diagnosis. Manjiviona (2003), for instance, has suggested that language and communication assessments undertaken with the ASD population should include investigations of pragmatics, verbal problem solving, and abstract language skills. To provide clinically useful information that can assist with the development of specific and relevant intervention goals, further analyses are, therefore, needed. Chapters 3, 4, and 5 will investigate the metalinguistic, pragmatic, and social problem solving skills in the children with a diagnosis of ASD.

48

3

Communicative Competence and Metalinguistic Ability in Children with Autism Spectrum Disorder 3

3.1

Introduction

Language and communication difficulties are central to the autistic presentation (APA, 1994). It is, therefore, not surprising that considerable research has investigated the language skills associated with AS and HFA. As presented in sections 2.1.1 and 2.1.2, validation and spectrum studies have focused on basic linguistic skills such as expressive, receptive, and/or syntactic skills (Klin et al., 1995; Prior et al., 1998; Szatmari et al., 1995; Szatmari et al., 1990). This perspective theoretically aligns with DSM-IV’s (APA, 1994) focus on language onset and early development. The study detailed in chapter 2 of this thesis extends previous findings by examining linguistic skills beyond expressive and receptive language and syntactic skills in the children with ASD. Chapter 2’s findings, however, may not adequately define the language and communication difficulties experienced in ASD. Metalinguistic language skills, for instance, emerge subsequent to the initial onset of language and include the ability to resolve complex and demanding language tasks such as resolving ambiguity (Sekerina et al., 2004) and producing and understanding figurative language (Levorato & Cacciari, 2002). These later-emerging language skills may likewise contribute to the many behavioural, social, and academic difficulties for children with a diagnosis of ASD (Lord, 2000).

3

“Communicative competence and metalinguistic ability: Performance by children and adults with Autism

Spectrum Disorder”, by F. M. Lewis, B. E. Murdoch, and G. C. Woodyatt, 2007, Journal of Autism and Developmental Disorders, 37, 1525-1538, was adapted from this chapter.

49 3.2


While the research focus remains on language skills that are relevant only in determining diagnosis, the wider clinical appreciation of the language skills in AS and HFA may be hindered. Diagnosis allows access to intervention (Bishop, 2006), but the basis for determining the level and type of educational and vocational support services should not be restricted to performance on functional linguistic tasks such as knowledge of vocabulary and grammatical competence only. More evidence on later-developing complex language skills is needed to direct clinical intervention with these client groups. 3.3


The current study aimed to broaden the scope of previous research into the language skills of individuals within the autism spectrum. Based on the theoretical framework of the validation studies (outlined in section 1.3.1), the first aim was to determine if performance on complex language tasks differed significantly when individuals with ASD were grouped according to developmental language onset as outlined in DSM-IV (APA, 1994). The second aim, based on the theoretical framework of the combined studies (as outlined in section 1.3.2), was to administer a language assessment comprised of later-emerging, complex, demanding language tasks to a group of children with a diagnosis of ASD (with no diagnostic differentiation undertaken) and typically developing children to determine if performance significantly differed between the two groups. The final aim, based on the theoretical framework of the spectrum studies to defining language characteristics (refer to section 1.3.3), was to examine the results of the language testing to determine if subgroups from the experimental group could be identified through hierarchical cluster analysis based on current performance on complex language tasks. Given previous conclusions of the differentiation of AS from HFA have been based on semantic/grammatical assessments only (e.g., Ghaziuddin et al., 2000; Howlin, 2003; Klin et al., 1995; Mayes & Calhoun, 2001; Szatmari et al., 1995; Szatmari et al., 1990), no hypothesis was offered regarding whether later-emerging complex language tasks differentiated the children grouped according to developmental language history. It

50 was hypothesised, however, that the ASD group would present with a significantly different profile when compared to their typically developing peers. As subgroups of individuals within the autism spectrum have been described previously, based on performance on basic functional language skills such as semantic and grammatical development (Prior et al., 1998), it was hypothesised that subgroups within ASD, based on performance on complex language tasks, would be evident from hierarchical cluster analysis. 3.4

Method

3.4.1

Participants

Recruitment details of the experimental and control participants involved in the present study have been presented previously (refer to section 2.4.1). Participant demographics are displayed in Table 2.1. 3.4.2

Measures

All participants were assessed using the Test of Language Competence-Expanded Edition (TLC-E) (Wiig & Secord, 1989) and the TONI-2 (Brown et al., 1990). The TONI-2 has been described in section 2.4.2. The TLC-E provided the language measures used to fulfil the three aims of the study. It is a standardised test for the paediatric population and assesses emerging metalinguistic competence in semantics, syntax, and/or pragmatics. The test utilises a strategy approach to language assessment based on the view that strategy acquisition is central to the growth in communicative competence and metalinguistic ability (Wiig & Secord). Level 2 of the TLC-E (Wiig & Secord, 1989) was used as it is suitable for children aged 9 years 0 months – 18 years 11 months. The children’s raw scores were converted to Standard Scores as per the instruction manual. The TLC-E subtests are Ambiguous Sentences, Listening Comprehension: Making Inferences, Oral Expression: Recreating Sentences, and Figurative Language. Examples of the four TLC-E subtests are presented in Appendix D.

51 The technical manual of the TLC-E Level 2 (Wiig & Secord, 1989) includes data on the psychometric properties of the test. Criterion-related validity, based on the correlations between the TLC-E Level 2 composite Standard Scores and the Test of Adolescent Language total test quotient (Hammill, Brown, Larsen, & Wiederholt, 1980) are moderately high for both language-learning disabled and nonlanguage-learning disabled adolescents (.74 and .62 respectively). The correlation of the TLC-E Level 2 composite with the Wechsler Intelligence Scale for Children-Revised (WICS-R) (Wechsler, 1974) Verbal Scale is notably higher than the correlation with the WISC-R Performance Scale, a scale designed to measure nonverbal intellectual performance, not verbal performance. The pattern of correlations supports the validity of the TLC-E Level 2 as a measure of language. The internal consistency reliability for Level 2 of the TLC-E is adequate (range: .75 - .82), as is the test-retest reliability (.78). 3.4.3

Procedure

The procedure using developmental language history for the reclassification of the ASD children into AS or HFA, as described in section 2.4.3, was undertaken for the current study. Table 2.1 displays the reclassified diagnoses. Standard Scores from all four TLC-E subtests (Wiig & Secord, 1989) were entered for the hierarchical cluster analysis. 3.5

Results

3.5.1

Performance on the TLC-E by the Children with ASD Reclassified as AS or HFA Based on Reported Developmental Language History

Group means and standard deviations for the reclassified AS and HFA groups are shown in Table 3.1. Comparative analyses of the performance on the TLC-E by the ASD children reclassified as AS or HFA were undertaken using Mann-Whitney U tests. For determination of statistical significance, a conservative level of p ≤ .05 was utilised. There were no significant differences detected between the two groups on age (p ≥ .05) and nonverbal intelligence (p ≥ .05). There was a significant difference in performance on Ambiguous Sentences (p ≤ .05), with the AS group outperforming the HFA group.

52 There were no significant differences between the reclassified groups on Listening Comprehension: Making Inferences, Oral Expression: Recreating Sentences, or Figurative Language (p ≥ .05 on all three measures).

Table 3.1 Descriptive statistics of the performance on the TLC-E by children with ASD reclassified as AS or HFA according to developmental language history Language onset

Language onset

Mann-Whitney U

prior to age two

after two years of

Test

years (AS)

age (HFA)

(n = 10)

(n = 8)

M (SD)

M (SD)

Z


Age (yrs, mths)

12, 5 (2, 4)

11, 4 (1, 11)

-1.022

.307

TONI-2

103 (18)

99 (20)

-0.867

.386

Amb Sent

7 (4)

4 (1)

-2.095

.036 *

List Comp

7 (3)

5 (2)

-1.572

.116

Oral Exp

5 (2)

5 (3)

-.184

.854

Fig Lge

8 (4)

6 (3)

-1.536

.125

TLC-E

Note. TLC-E = Test of Language Competence-Expanded Edition. ASD = Autism spectrum disorder. AS = Asperger syndrome. HFA = High functioning autism. TONI-2 = Test of Nonverbal Intelligence-2nd Edition. Amb Sent = Ambiguous Sentences. List Comp = Listening Comprehension: Making Inferences. Oral Exp = Oral Expression: Recreating Sentences. Fig Lge = Figurative Language. Yrs, mths = Years, months. * = p ≤ .05.

53

3.5.2

Performance by all Child Participants on the TLC-E

Group means and standard deviations for the measures from the study using a combined AS/HFA group of participants are shown in Table 3.2. For each measure, a Levene’s test was carried out, and where necessary, appropriate corrections for non-homogeneity of variance were undertaken. Student t tests indicated there was no significant difference between the two groups on age (p ≥ .05), but a significant difference on nonverbal ability (p ≤ .05).

Table 3.2

Descriptive statistics of the child participants on the TLC-E ASD

Control

(N = 20)

(N = 18)

t

Sig. (2tailed)

Variable

M

SD

M

SD

Age (yrs, mths)

11, 6

2, 2

11, 5

1, 8

0.223

.825

TONI-2

100

18

113

16

-2.259

.03 *

Amb Sent

5

3

10

3

-4.267

.001 ***

List Comp

6

3

8

2

-2.980

.005 **

Oral Exp

5

2

10

3

-6.699

.001 ***

Fig Lge

7

3

8

3

-1.856

.072

TLC-E

Note. TLC-E = Test of Language Competence-Expanded Edition. ASD = Autism spectrum disorder. TONI-2 = Test of Nonverbal Intelligence-2nd Edition. Amb Sent = Ambiguous Sentences. List Comp = Listening Comprehension: Making Inferences. Oral Exp = Oral Expression: Recreating Sentences. Fig Lge = Figurative Language. Yrs, mths = Years, months. * = p ≤ .05. ** = p ≤ .01. *** = p ≤ .001.

54 The children with ASD performed significantly less well than their normally developing peers on Ambiguous Sentences (p ≤ .001), Listening Comprehension: Making Inferences (p ≤ .01), and Oral Expression: Recreating Sentences (p ≤ .001). There was no significant difference between the two groups on Figurative Language (p ≥ .05). Descriptively, the standard deviations for the two groups are similar for Ambiguous Sentences and Figurative Language (refer to Table 3.2), suggesting a similar dispersion of scores for the ASD and control groups on these two TLC-E subtests. The standard deviation is greater for the ASD group on Listening Comprehension: Making Inferences, and smaller on Oral Expression: Recreating Sentences. Figures 3.1 – 3.4 show the dispersion of scores for the ASD and control groups. Two ASD children were identified as outliers on Ambiguous Sentences. Of the two participant groups, there is greater score dispersal for the ASD group on Listening Comprehension: Making Inferences and, despite similar standard deviations between the two groups, Figurative Language. On Ambiguous Sentences, there is a negative skew for the ASD group. In contrast, the Ambiguous Sentences median for the control group is centrally located. On Oral Expression: Recreating Sentences, 50 % of the ASD scores fall between Standard Scores of 3 and 6, with the remaining 50 % scoring above 6, while 50 % of the control group’s scores fall between 8 and 12, with the remaining 50 % distributed above and below these scores.

55 14 12

3, 4

10

Amb Sent

8

6

4 2

N=

20

18

ASD

Control GROUP

Figure 3.1.

Box plots displaying the dispersal of scores for the child ASD and

control groups’ performance on Ambiguous Sentences.

56

12

10

List

8

Comp 6

4

2

N=

20

18

ASD

Control GROUP

Figure 3.2.


control groups’ performance on Listening Comprehension: Making Inferences.

57

14 12

Oral Exp

10

8 6

4 2

N=

20

18

ASD

Control GROUP

Figure 3.3.


control groups’ performance on Oral Expression: Recreating Sentences.

58

16 14

12

Fig Lge

10

8 6

4 2

N=

20

18

ASD

Control GROUP

Figure 3.4.


control groups’ performance on Figurative Language.

3.5.3

Clusters of Children with ASD Based on Performance on the TLC-E

An agglomerative hierarchical cluster analysis was undertaken with the child ASD participants. Current performance on the four TLC-E subtests was used to determine an optimal classification. The hierarchical clustering of the ASD child participants is shown in Figure 3.5.

59

Figure 3.5.

Agglomerative hierarchical cluster analysis of the 20 ASD children

based on performance on the four subtests from the Test of Language Competence-Expanded Edition.

The cluster analysis identified three subgroups of children. These subgroups are descriptive only. Membership within the subgroups is shown in Table 3.3 (on following pages). Scores are given as within standard deviations from the control group means. Performance scores within each subgroup support the partitioning decision. Subgroup performance ranged from average to moderate-severe difficulties. Subgroup 3’s performance was generally average for Ambiguous Sentences, Listening Comprehension:

60 Making Inferences, and Figurative Language, but exhibited mild difficulties on Oral Expression: Recreating Sentences. Subgroup 2’s performance was within the average range for Listening Comprehension: Making Inferences and Figurative Language, and mild to moderately impaired on Ambiguous Sentences and Oral Expression: Recreating Sentences. Subgroup 1 was characterised by moderate-severe difficulties on Ambiguous Sentences, Listening Comprehension: Making Inferences, and Oral Expression: Recreating Sentences, and mild to moderate difficulties on Figurative Language. The Kruskal-Wallis test determined that age was not significantly different between the subgroups (χ2 = 2.838, p = .242). Performance on the TONI-2 was significantly different between the subgroups (χ2 = 8.461, p = .015). Post-hoc MannWhitney U tests identified a significant difference in the TONI-2 means between subgroups 1 and 2 (Z = -2.223, p = .026) and subgroups 1 and 3 (Z = -2.384, p = .013). There was no significant difference between Subgroups 2 and 3 (Z = -0.735, p = .556).

Table 3.3

Subgroups of ASD children based on performance on the TLC-E ID

Subgroup 1 (n = 11)

Given diag

Lge hist

Amb

List

Oral

Sent

Comp

Exp

Fig Lge

7

AS

Delayed

-2.5

-2.5

-2.5

-1.5


10

AS

Delayed

-2.5

-2.5

-2.5

-2.5

(SD: 1, 10)

19

ASD

Unsure

-2.5

-2.5

-2.5

-1.0

M TONI-2: 90 (SD: 12)

11

AS

Delayed

-2.0

-1.5

-2.5

-1.5

16

AS/ASD

Delayed

-2.5

-1.5

-2.5

-1.5

5

AS

Normal

-2.5

-2.0

-2.0

-2.0

12

AS

Unsure

-2.5

-2.0

-2.0

-1.5

1

AS

Normal

-2.0

-1.0

-2.0

-1.0

20

ASD

Normal

-2.0

-2.0

-2.5

-1.0

14

AS/ASD

Delayed

-2.5

-2.0

-1.5

-1.0

17

ASD

Delayed

-2.0

-2.5

-1.0

-2.0 Table continues

61

62


Subgroup 2 (n = 4)

Given diag

Lge hist

Amb

List

Oral

Sent

Comp

Exp

Fig Lge

6

AS

Normal

- 1.5

- 1.0

-1.5

- 1.0

15

AS

Normal

- 1.0

- 0.5

-2.0

- 1.0

8

AS

Normal

- 2.5

+0.5

-2.5

- 1.0

18

ASD

Delayed

- 2.0

- 0.5

-2.0

0.0

Subgroup 3 (n = 5)

2

ASD

Normal

+0.5

- 1.0

-2.0

+1.5

M age: 12, 8 (SD: 1, 8)

9

AS

Normal

- 1.0

- 2.0

-2.5

+1.0

M TONI-2: 100 (SD: 18)

3

AS

Normal

+1.0

+1.5

-1.0

+2.0

4

AS

Normal

+1.0

+1.0

-1.0

+0.5

13

AS

Delayed

- 1.0

+1.0

0.0

+1.0

M age: 12, 4 (SD: 3, 4) M TONI-2: 112 (SD: 15)

Table continues

Table 3.3 continued Note. TLC-E = Test of Language Competence-Expanded Edition. ASD = Autism spectrum disorder. AS = Asperger syndrome. ID = Identification number. Given diag = Independently determined diagnosis. Lge Hist = Developmental language history. Normal = Language onset prior to age two years. Delayed = Language onset after two years of age. Unsure = Language history not provided. Amb Sent = Ambiguous Sentences. List Comp = Listening Comprehension: Making Inferences. Oral Exp = Oral Expression: Recreating Sentences. Fig Lge = Figurative Language. TONI-2 = Test of Nonverbal Intelligence-2nd Edition. Yrs, mths = Years, months. Standard deviations above or below the control group means are shown. + = Standard deviations above the control group mean. - = Standard deviations below the control group mean.

63

64 3.6

Discussion

The major focus of the present study was to examine the performance on complex language tasks by a group of children with a diagnosis of ASD. While no prediction was made regarding performance of the individuals with ASD grouped according to developmental language history, the current findings suggest that developmental language history may not be a significant factor in current performance on complex language tasks. The findings, however, support the study’s two predicted hypotheses. Firstly, performance on a range of complex language tasks that emerge later in the developmental process of language acquisition was less proficient in ASD than in typically developing peers. Secondly, subgroups within ASD could be described, based on current performance on later-developing complex language tasks. 3.6.1

Metalinguistic Abilities in Children with ASD Reclassified as AS or HFA According to Reported Developmental Language History

The findings of the current study suggest that, of the later-developing complex language skills, performance on resolving ambiguity at the word and sentence level differentiated the ASD children when reclassified as AS or HFA using the language criterion from DSM-IV (APA, 1994). However, as discussed in section 2.6.1, the results should be interpreted cautiously. One concern raised by previous researchers (Klin et al., 2005) is that the DSMIV (APA, 1994) distinction between AS and HFA may not consistently be applied in diagnostic and clinical practice. As noted previously (refer to section 2.6.1), there are examples in the current study of children not using single words by two years of age and possibly not using communicative phrases by three years of age. According to DSM-IV criteria, these children would be given the diagnosis of AD. However, a number of children in the present study, with these milestones, were given a diagnosis of AS. Another concern regarding the validity of diagnostic decisions based on developmental milestones is that the interpretation of reported milestones may be inconsistent, leading to conflicting diagnoses (Tager-Flusberg, 2003; WoodburySmith et al., 2005). Tager-Flusberg, for instance, argued that DSM-IV’s language criterion would be judged as delayed onset if compared to normative data from

65 typically developing children. In such a presentation, a diagnosis of AD, not AS, could be given. The methodology applied in the present study avoided these two issues. The children in this study were reclassified by two experienced speech pathologists in strict accordance to DSM-IV’s (APA, 1994) language criterion. Those with an onset of single words prior to two years of age were reclassified as AS, and those whose onset of single words was after two years of age were reclassified as HFA. This methodology, however, cannot avoid the issue of diagnostic reclassification based on possibly unreliable retrospective recall of developmental milestones (Tager-Flusberg, 2003; Woodbury-Smith et al., 2005). For this reason, although developmental language history was not a differentiating factor in all but one language measure, interpreting the findings as suggesting that complex language skills do not differentiate individuals with ASD with differing developmental language histories needs to be considered with caution. 3.6.2

Metalinguistic Language Abilities in Children with ASD

The language skills assessed by the TLC-E (Wiig & Secord, 1989) require high-level cognitive-linguistic skill. The children with ASD investigated in the study experienced difficulties with the complex language tasks relative to their peers. They were less skilled on tasks of resolving ambiguity, understanding inferential language, and using linguistic flexibility to produce speech acts constrained by a communicative situation. Cognitive skill may have contributed to these difficulties, as the nonverbal cognitive skills of the children with ASD were significantly below the normally developing control peers. Previous studies of autism spectrum presentations have also identified difficulties with later-developing language tasks. For example, difficulties with inferential language (Dennis, Lazenby, & Lockyer, 2001; Koning & Magill-Evans, 2001; Ozonoff & Miller, 1996), figurative language (Dennis et al.), and producing relevant speech acts (Dennis et al.; Losh & Capps, 2003; Ozonoff & Miller) have all been described in the AS/HFA population. The current findings support these claims. Some inconsistencies are, however, apparent. Contrary to the present findings and

66 those of Dennis et al., Gunter et al. (2002) described inferential skills in AS/HFA as being no different to controls. Further, Dennis et al. described children younger than the experimental group in the current study as possessing some level of linguistic flexibility for resolving ambiguity. Heterogeneity within the ASD group of children was noted on some of the metalinguistic tasks investigated in the present study, as was the case for performance on all the language measures assessed by the CELF-4 (Semel et al., 2003) in section 2.5.2. This level of analysis, where heterogeneity is noted but not further investigated, offers limited clinical utility as individual differences may require an individualised approach to intervention planning. 3.6.3

Heterogeneity of Metalinguistic Skills in Children with ASD

Extending the analyses to an examination of subgroups within ASD based on current performance on complex cognitive-linguistic tasks, as assessed by the TLC-E (Wiig & Secord, 1989), provided research findings with additional clinical relevance. The children on the autism spectrum exhibited a range of abilities not evident when viewed as a single experimental group, as presented in section 3.6.2. Nonverbal cognitive skill was significant in the subgroup formation. This finding supports the findings (see section 2.6.3) of a possible relationship between cognitive skill and language competence. Metalinguistic ability within the ASD subgroups ranged from average to above average through to moderate-severe impairments. Children from the most competent subgroup (subgroup 3) were generally able to resolve ambiguity. Most of these participants were also able to make correct inferential judgments. They showed some linguistic flexibility by including all three words into a context-dependent sentence as required by the Oral Expression: Recreating Sentences subtest. However, their responses were often marked by syntactic errors. For instance, one child from subgroup 3 produced Before we go we’ll rather see this movie after we get some popcorn (target words are in bold). Children from the most competent subgroup, while able to generally use the context to derive the meaning of figurative speech, exhibited restricted knowledge of everyday figures of speech.

67 Performance by the least competent subgroup (subgroup 1) was marked by an inability to recognise ambiguity. Most of the children in this subgroup gave either one response or elected to pass on the Ambiguous Sentence items. Some children from the subgroup, however, gave two responses, without realising the two responses did not provide two differing interpretations. For instance, for the item “I saw the girl take his picture”, one child’s response was to take the picture from him and took his pictures. This particular child missed the ambiguity of the words “take” and “picture”. Generally, participants from the least competent subgroup were inconsistent in selecting correct responses in the Listening Comprehension: Making Inferences subtest. The inability to consistently select even one correct response could indicate a random selection process rather than an ability to detect inferential messages. The ASD children in the least competent subgroup exhibited difficulty with the linguistic and pragmatic constraints of the Oral Expression: Recreating Sentences tasks. One child produced This bakery is fresh said the black lady. The other lady said there’s no way I’m not buying here, despite being requested to use the three words (fresh, nor, here) in one sentence only. Children from this subgroup were also, generally, not able to provide their own meaning for the figurative language tasks, nor were they able to derive meaning from the context. Most selections were literal interpretations of the target. For example, for the item “A student after listening to her friend says I just can’t swallow that”, one ASD child provided not edible as a (literal) interpretation in his own words. When given a choice of four interpretations to select from, he chose I don’t want to chew on it. As a non-delineated group, the children with ASD were identified as experiencing difficulties with resolving ambiguity, understanding inferential language, using linguistic flexibility to produce speech acts constrained by a communicative situation, but not with interpreting figurative language. Viewing ASD as a dimensional disorder on an autism spectrum, however, identified specific difficulties, such as with figurative language, which can be incorporated into intervention planning for individual children.

68 3.7

Conclusion

Using similar theoretical perspectives but broadening the methodologies of previous research into the language skills in AS and HFA, the results from the current study extend the clinical appreciation of the language difficulties associated with ASD. Although developmental language history was found to be significant in the outcome of skill in resolving ambiguity for the children reclassified as AS or HFA using DSMIV (APA, 1994) criteria, the reliability of the findings can be challenged due to the reliance upon retrospective recall of developmental milestones. The findings suggest that the language difficulties experienced by children with ASD extend beyond semantic and syntactic skills as defined by previous studies. Viewed as a combined ASD group, the children investigated in the study presented with difficulties on a range of metalinguistic tasks. More clinically relevant is the identification of subgroups within ASD, defined by their complex language skills. The clinical implications of the findings suggest that conceptualising AS and HFA as autistic disorders on a spectrum may provide much clinically relevant information about the range of language skills within the autistic presentation. The reliance on accurate assessment of current language performance, rather than a reliance on possibly inaccurate recall of past language performance should form the basis for intervention strategies and services. The findings of the current study suggest that to clinically appreciate the diversity of language skills in ASD, comprehensive language assessments should be undertaken on all individuals. Individualised education and vocational support plans can then be determined, based on individual strengths and weaknesses. Chapters 2 and 3 have focused on conventional language skills. Manjiviona (2003), however, has suggested a comprehensive language and communication assessment should include an assessment of social communication skills. Chapter 4 examines social communication or pragmatic language skills in the children with ASD.

69

4

Pragmatic Language Skills in Children with Autism Spectrum Disorder

4.1

Introduction

Bloom and Lahey (1978) conceptualised the development of language as including three major components. Language content reflects semantic development, language form reflects syntactic and morphologic development, and language use reflects pragmatic language, or the way language is used in a social context. Pragmatic difficulties may be impediments to full and rewarding participation in academic and social activities (Martin & McDonald, 2004). Manjiviona's (2003) proposal that an assessment of pragmatic skills should be included in the diagnostic test battery for children with AS, HFA, or ASD appears, therefore, to be clinically justified. Language content and form are easily evaluated using traditional assessments of language such as the CELF-4 (Semel et al., 2003). Such assessments, however, provide limited information on pragmatic communication in social situations where interpretation of a message is derived from the communicative context (Bishop, 1997, 2003). Pragmatic skills require an individual to use situational and communicative contexts, as well as understand the speaker’s intentions (Sabbagh, 1999). A range of verbal and nonverbal communicative skills come under the banner of pragmatic language skills. These include appreciating humour and indirect requests, understanding metaphorical and inferential language, discourse skills (Ozonoff & Miller, 1996; Sabbagh), eye contact (Tantam, Holmes, & Cordess, 1993), and the use and comprehension of prosody (McCann & Peppe, 2003). While pragmatic difficulties may be one aspect of a non-autistic child’s communicative difficulties (Rapin & Allen, 1983), they are central to the communication difficulties associated with both AS and AD (APA, 2000). It is not surprising, therefore, that pragmatic language skills in AS and AD have been the

70 focus of much research (e.g., Dennis et al., 2001; Emerich et al., 2003; Fine et al., 1991; Fine et al., 1994; Ghaziuddin & Gerstein, 1996; Gilchrist et al., 2001; Gunter et al., 2002; Paul et al., 2005; Ramberg et al., 1996; Shriberg et al., 2001; Verte, Geurts, Roeyers, Rosseel, et al., 2006; Volden, 2004). All three theoretical approaches outlined in section 1.3 have been used to examine pragmatic skills in AS and/or HFA. The clinical impression of the pragmatic profile associated with AS, HFA, or ASD gained from the research findings to date, however, is unclear, due to differing methodologies, different pragmatic tasks under investigation, differently defined participant groups, an absence of replication studies to substantiate the findings, and/or a failure to fully appreciate the conceptualisation supporting the theoretical framework applied. 4.1.1

Validation Approach to Investigating Pragmatic Skills in AS and HFA

A number of researchers have adopted the theoretical approach of validation studies, and examined pragmatic skills, reciprocity in conversations, discourse skills, and prosodic characteristics of AS and HFA (see e.g., Fine et al., 1991; Fine et al., 1994; Ghaziuddin & Gerstein, 1996; Gilchrist et al., 2001; Ramberg et al., 1996; Shriberg et al., 2001; Verte, Geurts, Roeyers, Rosseel, et al., 2006). The research findings, however, have failed to consistently support the validation of AS as a separate diagnosis. Ramberg et al. (1996), for instance, examined a range of verbal and nonverbal pragmatic skills in AS and HFA. They found no difference between the two disorders on conversational structure, nonverbal communication (response to eye contact and use of gesture), and intonation patterns in speech. Verte, Geurts, Roeyers, Rosseel, et al. (2006), moreover, described no differences between AS and HFA on a comprehensive range of pragmatic functions. They administered the Children’s Communication Checklist (Bishop, 1998) to groups of children assigned a diagnosis of AS, HFA, or pervasive developmental disorder-not otherwise specified on the basis of modified DSM-IV criteria (APA, 1994). Diagnoses were validated using the Autism Diagnostic Interview-Revised (ADI-R) (Lord, Rutter, & LeCouteur, 1994). Using parental ratings, Verte and colleagues found that inappropriate initiation of interactions, coherence, stereotypical conversations, use of conversational context,

71 and conversational rapport failed to differentiate the children with a clinical diagnosis of AS from those with a diagnosis of HFA. In contrast, other researchers have suggested that the two disorders can be differentiated on pragmatic tasks. Individuals with AS have been described as better conversationalists than those with HFA (Gilchrist et al., 2001). The failure to build on previous exchanges resulting in a difficulty in building reciprocal conversation, for instance, has been identified as characterising individuals with HFA (Fine et al., 1994). Individuals with AS were reported to have a tendency for obsessive and repetitive topic expression (Shriberg et al., 2001) and higher total pedantic speech scores (Ghaziuddin & Gerstein, 1996) relative to those with HFA. Discourse skills associated with AS have been described as more complex than those with HFA (Fine et al.), with individuals with AS using more bridging devices. Individuals with AS, however, are reported to be more likely than those with HFA to make errors in the use of cohesive links (Fine et al.). Moreover, speech and prosodic characteristics may differentiate the two disorders (Shriberg et al.), with individuals with a diagnosis of AS reportedly using more functionally useful intonation and less non-useful intonation patterns than those with a diagnosis of HFA (Fine et al., 1991). Any differences identified between AS and HFA, however, appear to be quantitative, not qualitative (see e.g., Szatmari et al., 1995). In conclusion, a range of verbal and nonverbal pragmatic skills have been investigated in AS and HFA through validation studies. Investigations have included the use of eye contact, prosody, and initiation, reciprocity and cohesion in conversations. There is, however, no consensus as to which pragmatic skills, if any, differentiate AS from HFA, due in part, to the different pragmatic skills assessed by individual research groups. 4.1.2

Combined Approach to Investigating Pragmatic Skills in AS and HFA

Few studies (Dennis et al., 2001; Emerich et al., 2003) have investigated pragmatic language skills where AS and HFA have been combined into a single experimental group. Emerich et al., using parental report of diagnosis, investigated the comprehension of humorous material in an undifferentiated AS/HFA group of

72 adolescents. The AS/HFA group was matched to typically developing children on age and gender. Verbal ability did not differ between the two groups. Emerich and colleagues found the comprehension of jokes to be poorer in the AS/HFA group relative to the control group. Dennis et al.’s (2001) study focused on non-inferential (linguistic) and inferential (pragmatic) language skills in a small (N = 8) group of AS/AD children (combined to maximise the sample size). All AS/AD children had verbal IQs greater than 70. The non-inferential tasks administered included oral vocabulary and ambiguity, and the inferential or pragmatic tasks included presupposition and implication, script inferencing tasks such as the determination of logical coherence of predictable social scripts, metaphor tasks, and speech act tasks reflecting inferencing about intentions. Control data came from three different control groups, with each control group matched to the AS/AD group on chronological age and gender. There were no significant performance differences between the AS/AD group and the control groups assessed using non-inferential tasks of ambiguity and oral vocabulary. The AS/AD group, however, was significantly less competent than the control groups on the pragmatic tasks of making pragmatic inferences, understanding metaphors, and understanding intentionality in social contexts. In conclusion, relatively few studies have examined pragmatic skills in combined AS/HFA groups of participants. It is difficult to make comparisons between the findings of the studies, as different participant inclusion criteria for diagnosis were employed. Emerich et al. (2003) for instance, relied upon parental reporting of diagnosis, while Dennis et al.’s (2001) inclusion criteria were more stringent, requiring a DSM-IV (APA, 1994) diagnosis of AS or AD on the basis of a two hour interview and a quantitative rating on the Childhood Autism Rating Scale (Schopler, Reichler, & Renner, 1988). Nonetheless, difficulties with humour (Emerich et al.), metaphorical and inferential language, and understanding intentionality in social contexts (Dennis et al.) have been described in experimental groups where children with AS and/or HFA have been combined into non-differentiated groups. Replication studies, however, are needed to substantiate these findings.

73 4.1.3

AS-specific Studies Investigating Pragmatic Skills

One known study (Gunter et al., 2002) has focused on pragmatic skills in a group of (predominantly) children and adolescents with AS matched on age and verbal IQ to non-neurologically impaired controls. Verbal skills were assessed using the Right Hemisphere Language Battery (Bryan, 1989) and the Unusual Metaphors Test (Bottini et al., 1994). Gunter et al. examined competence in appreciating humour, interpreting well-known metaphors presented in a picture mode and verbal/written mode, unusual metaphors, and inferential language. Gunter et al.’s finding of difficulties with humour in AS is consistent with Emerich and colleagues (2003), who described similar difficulties in their AS/HFA sample. Gunter et al., however, described intact metaphorical and inferential language in AS, which is at odds with Dennis et al.’s (2001) earlier findings (see section 4.1.2 above). In addition to appreciation of humour, only proficiency in interpreting unusual metaphors differentiated Gunter et al.’s AS group from the control group. Making comparisons between the findings from Gunter et al. (2002), Dennis et al. (2001), and Emerich et al. (2003) fails to extend the clinical understanding of pragmatic difficulties associated with AS and or AS/HFA, as methodological differences in participant inclusion criteria may account for the inconsistent results between findings. Comparing Gunter et al.’s findings with AS-specific groups of participants, such as in the validation studies, likewise fails to extend the clinical understanding of pragmatic difficulties associated with AS as no validation study has examined skills such as humour appreciation and metaphorical and inferential language. 4.1.4

Spectrum Approach to Investigating Pragmatic Skills in ASD

A small number of studies have adopted the theoretical position that AS and HFA are similar presentations on a spectrum of autistic disorders. As with the studies presented in the preceding sections (sections 4.1.1, 4.1.2, 4.1.3), the different skills investigated in the studies make it difficult to understand accurately the clinical presentation of pragmatic skills associated with ASD. Paul et al. (2005), for instance, investigated children with verbal IQs greater than 70 and aged between 14 – 21 years with any

74 form of ASD matched to typically developing control children. Diagnoses of ASD were validated by the ADI-R (Lord et al., 1994) and the ADOS-G (Lord et al., 2000). Paul et al. described significant between-group differences in aspects of prosodic perception and production. In contrast, Volden (2004) focused on conversational repair in children with ASD matched to generally younger children with similar language ages. Volden described the two groups as having similar skills in responding to requests for clarification and the use of a variety of repair strategies. The ASD group, however, was significantly more likely to respond inappropriately to requests for clarification. Although the conceptualisation of ASD as a spectrum disorder suggests that a range of competencies may be evident within the presentation (Wing, 1989), neither Paul et al. (2005) nor Volden (2004) undertook within-group analyses to investigate the range of pragmatic skills within their participant groups. Research studies investigating the range of skill within ASD are needed to clarify the clinical needs of children with ASD. 4.2


The debate regarding the identification of subtypes within pervasive developmental disorders such as AS and HFA is diagnostically relevant (see e.g., Pomeroy, 1998), as the given diagnostic label will determine the type of information, support, and resources offered (American Speech-Language-Hearing Association, 2006). Clinical practice, however, requires clinically applicable methods of classifying and conveying information (Wing, 1998) for the planning of intervention services. An accurate understanding of specific pragmatic difficulties experienced by children with ASD is necessary before support can be implemented to enhance participation and success in academic and social activities. To gain a clearer clinical picture of the pragmatic profile in AS, HFA, and/or ASD, it may be necessary for studies to assess a range of verbal and nonverbal pragmatic skills in one sample of participants, followed by replication studies to substantiate the findings.

75 4.3


The current study aimed to investigate verbal and nonverbal pragmatic skills in children with ASD by analysing data from an objective pragmatic assessment and subjective pragmatic ratings. As in the validation studies presented in section 1.3.1, a comparative study of pragmatic functions was undertaken, using DSM-IV’s (APA, 1994) language criterion to reclassify participants as AS or HFA. Additionally, the present study adopted the theoretical perspective of viewing AS and HFA as similar autistic presentations (as in section 1.3.2), and investigated the proficiency of the children in a single experimental ASD group relative to typically developing peers. Finally, adopting a spectrum approach (as discussed in section 1.3.3), two hierarchical cluster analyses of the ASD participants were undertaken to examine subgroups within the ASD group of children based on performance on an objective assessment of pragmatic skills and subjective ratings of pragmatic skills. No hypotheses were made regarding the analyses using the subjective ratings of pragmatic skills. Due to inconsistent findings (as presented in section 4.1.1) of any categorical distinction between AS and HFA on a range of verbal and nonverbal pragmatic tasks, no hypothesis was made regarding the outcome of an examination of the reclassification study. Given the inconsistent findings in section 4.1.2 of pragmatic language deficits in combined AS/HFA studies, no hypothesis was made regarding the comparison between the undifferentiated ASD group of children and the control children. As there is evidence of subgroups within ASD based on language performance, for instance from Prior et al. (1998) and from the discussion in section 2.6.3, it was hypothesised that subgroups of children, differing on pragmatic skills, would be evident within the ASD group of children. 4.4

Method

4.4.1

Participants

See section 2.4.1 for participant and recruitment details. Participant demographics are displayed in Table 2.1.

76 4.4.2

Measures

All participants were assessed using the TONI-2 (Brown et al., 1990), the Right Hemisphere Language Battery (RHLB) (Bryan, 1989), and the Children’s Communication Checklist-Second Edition (CCC-2) (Bishop, 2003). Information regarding the TONI-2 and its inclusion in the test battery is presented in section 2.4.2. The RHLB was designed to provide quantitative and qualitative information of language impairment following right hemisphere damage in adults (Bryan, 1989), but has previously been used in research with children and adolescents with AS (e.g., Gunter et al., 2002). It is acknowledged that individuals with ASD do not have localised right hemisphere damage, nevertheless, the RHLB was chosen as it is one of the few objective assessments available that examines the use of contextual information to derive meaning. The RHLB consists of seven subtests, six of which were used in the current study. These were Metaphor Picture Test, Written Metaphor Test, Comprehension of Inferred Meaning, Appreciation of Humour, Lexical Semantic Test, and Production of Emphatic Stress. The seventh subtest, Discourse Analysis, was not administered for the study due to time and financial constraints. Appendix E provides examples of each of the six subtests. Limited data on the psychometric properties of the RHLB (Bryan, 1989) are contained in the test manual. The manual cites references where psychometric properties of the RHLB are reported, but the cited references provide limited information or are not accessible (e.g., conference proceedings). The CCC-2 (Bishop, 2003) provides subjective data on a range of pragmatic measures. It consists of 70 multiple choice items completed by an adult familiar with the child. The items are divided into 10 scales. The scales are Speech, Syntax, Semantics, Coherence, Inappropriate Initiation, Stereotyped Language, Use of Context, Nonverbal Communication, Social Relations, and Interests. Response choices are 0 (less than once a week or never), 1 (at least once a week, but not every day), 2 (once or twice a day), or 3 (several times [more than twice] a day [or always]). The CCC-2 manual contains tables to convert raw scores to age-adjusted scaled scores. Two composite scores from the CCC-2 can be calculated, however, Bishop

77 advises that the use of scaled scores is appropriate in research settings. The CCC-2 testing manual (Bishop, 2003) contains psychometric measures regarding the reliability and validity of the screening tool. The internal consistency of the CCC-2 is adequate, with the internal consistency Alpha values for each scale ranging from .65-.80. Correlation coefficients for inter-rater reliability (parent rating and teacher rating) range from .15-.79. Inter-rater reliability for Stereotyped Language and Use of Context did not reach statistical significance. Bishop argued the low interrater agreement on pragmatic functions was not unexpected given pragmatic function is context dependent, and different raters observe children in different contexts. Validation data for the CCC-2 (Bishop, 2003) were collected from three clinical samples (children with a range of communicative problems, children on speech pathology caseloads, and children referred to a tertiary developmental paediatric clinic). In general, the CCC-2 yielded profiles that were consistent with clinical expectations, confirming the utility of the checklist as a screening instrument (Bishop). 4.4.3

Procedure

Three levels of analyses, reflecting the three theoretical approaches to research identified in the literature review (refer to section 1.3), were undertaken for the current study. For each level of analysis, an analysis of the objective measures from the RHLB (Bryan, 1989) and an analysis of the subjective ratings from the CCC-2 (Bishop, 2003) was undertaken. Completed CCC-2 (Bishop, 2003) forms were returned for a total of 35 children (19 ASD participants and 16 control participants). For each child, two independent informants familiar with the child completed the CCC-2 forms. For the ASD group of participants, the first informants (Informants Number 1) included mothers, fathers, and grandparents. The second informants (Informants Number 2) for the ASD children included mothers, fathers, family friends, teachers, grandparents, and support workers. Informants Number 1 for the control children included mothers, fathers, and grandparents. Mothers, fathers, grandparents, family friends, step-parents, and teachers formed the control group’s Informants Number 2. The participants’ two

78 informants were requested to complete the CCC-2 independently without collaboration. Pearson inter-rater reliability coefficients were calculated for the two informant groups for the 35 children. Inter-rater reliability for each of the 10 CCC-2 scales was adequate (see Table 4.1).

Table 4.1

Pearson correlations between ratings by

Informants 1 and Informants 2 on the 10 CCC-2 scales for the 35 children CCC-2 Scale

Pearson r

Significance (2-tailed)

Speech

.634

.001 **

Syntax

.627

.001 **

Semantics

.712

.001 **

Coherence

.794

.001 **

Inappropriate Initiation

.761

.001 **

Stereotyped Language

.815

.001 **

Use of Context

.865

.001 **

Nonverbal Communication .874

.001 **

Social Relations

.944

.001 **

Interests

.716

.001 **

Note. CCC-2 = Children’s Communication Checklist-2nd Edition. ** = Correlation is significant at the 0.01 level.

Ten scaled scores and two composite scores were calculated for the 35 participants. For the purpose of the present study, data provided by Informants Number 1 on the Coherence, Inappropriate Initiation, Stereotyped Language, Use of Context, and Nonverbal Communication scales of the CCC-2 (Bishop, 2003) only

79 were used for the statistical analyses. Appendix F gives an example question for each of the CCC-2 scales used in the study. The reclassification of children with a diagnosis of ASD is outlined in section 2.4.3, with the reclassified diagnoses shown in Table 2.1. Raw RHLB (Bryan, 1989) scores were entered for the statistical analyses as the test is not standardised for the paediatric population. Two cluster analyses were undertaken on the ASD data. Performance scores from all six RHLB subtests were used for the first analysis and the data from the five CCC-2 scales (Bishop, 2003) were entered for the second analysis. 4.5

Results

4.5.1

Examination of Pragmatic Performance by the Children with ASD Reclassified as AS or HFA Based on Reported Developmental Language History

4.5.1.1 Objective pragmatic assessment using the RHLB. Group means and standard deviations for all objective measures are shown in Table 4.2. For determination of statistical significance, a conservative level of p ≤ .05 was utilised. Non-parametric Mann-Whitney U tests indicated that there were no significant differences between the two reclassified groups (AS, HFA) of children on age or performance on the TONI-2 (p ≥ .05). Performance on Metaphor Picture Test (p ≤ .05) was significantly different; with the AS group outperforming the HFA group. Performance on the remaining RHLB measures failed to reach significance (p ≥ .05 on Written Metaphor Test, Comprehension of Inferred Meaning, Appreciation of Humour, Lexical Semantic Test, and Production of Emphatic Stress).

80

Table 4.2 Descriptive statistics of the performance on the RHLB by children with ASD reclassified as AS or HFA according to developmental language history Variable

Language onset

Language onset

Mann-Whitney U

prior to two

after two years

Test

years of age

of age

(AS)

(HFA)

(n = 10)

(n = 8)

M (SD)

M (SD)

Z


Age (yrs, mths)

12, 5 (2, 4)

11, (1, 11)

-1.022

.307

TONI-2

103 (18)

99 (20)

-0.867

.386

M P (10 a)

8 (2)

6 (2)

-2.162

.031 *

a

W M (10 )

9 (2)

7 (3)

-1.763

.078

a

7 (3)

4 (3)

-1.581

.114

6 (2)

4 (3)

-1.757

.079

L S (20 )

20 (0)

20 (0)

-0.169

.866

P E S (10 a)

8 (2)

7 (2)

-0.678

.498

RHLB

C I M (12 ) a

A H (10 ) a

Note. RHLB = Right Hemisphere Language Battery. ASD = Autism spectrum disorder. AS = Asperger syndrome. HFA = High functioning autism. TONI-2 = Test of Nonverbal Intelligence-2nd Edition. M P = Metaphor Picture Test. W M = Written Metaphor Test. C I M = Comprehension of Inferred Meaning. A H = Appreciation of Humour. L S = Lexical Semantic Test. P E S = Production of Emphatic Stress. Yrs, mths = Years, months. a

= Maximum possible score.

* = p ≤ .05.

81 4.5.1.2 Subjective pragmatic assessment using the CCC-2 ratings. Group means and standard deviations for all subjective measures from the CCC-2 scales are shown in Table 4.3. Mann-Whitney U tests indicated that there were no significant differences between the reclassified groups on any of the five CCC-2 scale means (p ≥ .05 on all measures).

Table 4.3

Descriptive statistics of Informants Number 1 ratings on the CCC-

2 for children with ASD reclassified as AS or HFA according to developmental language history

CCC-2 Scale

Language onset

Language

Mann-Whitney U

prior to two years

onset after two

Test

of age

years of age

(AS)

(HFA)

(n = 9) a

(n = 8) a

M (SD)

M (SD)

Z


Coherence

3.3 (1.1)

3.2 (2.4)

-.344

.731

Inappropriate Init 4.0 (2.0)

3.4 (2.4)

-.585

.558

Stereotyped Lge

2.9 (1.6)

3.7 (3.3)

-.098

.992

Use of Context

2.1 (1.4)

2.7 (4.1)

-.548

.583

Nonverbal Com

1.9 (1.4)

3.1 (2.5)

-1.125

.261

Note. CCC-2 = Children’s Communication Checklist-2nd Edition. ASD = Autism spectrum disorder. AS = Asperger syndrome. HFA = High functioning autism. Inappropriate Init = Inappropriate Initiation. Stereotyped Lge = Stereotyped Language. Nonverbal Com = Nonverbal Communication. a

= Reduced numbers due to missing data (CCC-2 forms not returned).

82

4.5.2

Examination of Pragmatic Performance by all Child Participants

4.5.2.1 Objective pragmatic assessment using the RHLB. Group means and standard deviations for all objective pragmatic measures are shown in Table 4.4. There was no significant difference between the ASD group and the control group on age (p ≥ .05), but a significant difference on nonverbal intelligence (p ≤ .05). For each measure, a Levene’s test was carried out, and where necessary, appropriate corrections for non-homogeneity of variance were undertaken. Student t tests detected significant differences in the means between the two groups on three RHLB measures. The ASD group was less proficient on Lexical Semantic Test (p ≤ .05), Comprehension of Inferred Meaning (p ≤ .01), and Production of Emphatic Stress (p ≤ .01). All other measures were not significant (p ≥ .05).

83 Table 4.4 Descriptive statistics of the child participants on the RHLB ASD

Control

(N = 20)

(N = 18)

t

Sig. (2tailed)

Variable

M

SD

M

SD

Age (yrs, mths)

11, 6

2, 2

11, 5

1, 8

0.223

.825

TONI-2

100

18

113

16

-2.259

.03 *

M P (10 a)

6.6

3.3

7.6

2.8

-0.953

.347

a

W M (10 )

7.3

3.2

8.4

2.3

-1.235

.225

a

C I M (12 )

5.7

2.8

8.1

2.2

-2.837

.007 **

A H (10 a)

4.9

2.9

6.1

1.8

-1.500

.143

L S (20 a)

19.5

0.7

19.9

0.2

-2.208

.037 *

P E S (10 a)

7.4

2.0

9.2

1.0

-3.348

.002 **

RHLB

Note. RHLB = Right Hemisphere Language Battery. ASD = Autism spectrum disorder. TONI-2 = Test of Nonverbal Intelligence-2nd Edition. M P = Metaphor Picture Test. W M = Written Metaphor Test. C I M = Comprehension of Inferred Meaning. A H = Appreciation of Humour. L S = Lexical Semantic Test. P E S = Production of Emphatic Stress. Yrs, mths = Years, months. a

= Maximum possible score.

* = p ≤ .05. ** = p ≤ .01.

Descriptively, the standard deviations for performance on all six RHLB subtests are greater for the ASD group than the control group suggesting greater score dispersal in the ASD group (refer to Table 4.4). The dispersal of scores for the two experimental groups on the six RHLB tests is displayed in Figures 4.1 – 4.6. There were no outliers in the ASD group on any of the measures. Outliers and/or far outliers from the control group were noted on Metaphor Picture Test, Written Metaphor Test, Lexical Semantic Test, and Production of Emphatic Stress. Of the two groups, the box plot is longer for the ASD group compared to the control group on all six RHLB

84 measures, and the middle 50 % of ASD scores on all RHLB measures except for Comprehension of Inferred Meaning extends further than the control group’s scores, both indicating greater score dispersal for the ASD group on five of the six RHLB measures.

12 10

8 6

15

MP 4 2

14

0

4 17

-2

N=

18

20

Control

ASD GROUP

Figure 4.1.


and control groups’ performance on Metaphor Picture Test.

85

12 10

8

WM

6

4 2

0

17 -2

N=

20

ASD

Figure 4.2.

18

GROUP

Control


and control groups’ performance on Written Metaphor Test.

86

12

10

8

CIM

6

4 2

0 -2

N=

20

ASD

Figure 4.3.

18

GROUP

Control


and control groups’ performance on Comprehension of Inferred Meaning.

87

10

8

AH

6

4

2

0

N=

20

ASD

Figure 4.4.

18

GROUP

Control


and control groups’ performance on Appreciation of Humour.

88

20.5

20.0

19.5

LS 19.0

15 18.5

18.0

17.5

N=

20

18

ASD GROUP

Figure 4.5.

Control


and control groups’ performance on Lexical Semantic Test.

89

12

10

PES

8

12

6

17

4

2

N=

20

ASD

Figure 4.6.

18

Control GROUP


and control groups’ performance on Production of Emphatic Stress.

4.5.2.2 Subjective pragmatic assessment using the CCC-2 ratings. Means and standard deviations for all subjective measures from the CCC-2 are shown in Table 4.5. For each measure, a Levene’s test was carried out, and where necessary, appropriate corrections for non-homogeneity of variance were undertaken. Student t tests detected significant differences on all CCC-2 scales. The mean scaled score for the ASD group was significantly lower than controls on Coherence (p ≤ .001), Inappropriate Initiation (p ≤ .001), Stereotyped Language (p ≤ .001), Use of Context (p ≤ .001), and Nonverbal Communication (p ≤ .001).

90

Table 4.5 Descriptive statistics of the child participants on the CCC-2 scales based on ratings by Informants Number 1 ASD (N = 19) CCC-2 Scale

M

Control a

SD

(N = 16) a M

SD

t

Sig. (2tailed)

Coherence

3.2

1.7

10.4

2.5

-9.678

.001 ***

Inapprop Init

3.7

2.1

9.7

2.8

-7.382

.001 ***

Stereotyped Lge

3.6

2.5

10.4

2.3

-8.324

.001 ***

Use of Context

2.4

2.8

9.9

3.0

-7.647

.001 ***

Nonverbal Com

2.7

2.0

10.5

1.9

-11.691

.001 ***

Note. CCC-2 = Children’s Communication Checklist-2nd Edition. ASD = Autism spectrum disorder. Inapprop Init = Inappropriate Initiation. Stereotyped Lge = Stereotyped Language. Nonverbal Com = Nonverbal Communication. a

= Reduced numbers due to missing data (CCC-2 forms not returned).

*** = p ≤ .001.

Descriptively, the standard deviations for the ASD group on Coherence, Inappropriate Initiation, and Use of Context are smaller than the control group (refer to Table 4.5) suggesting less score dispersal on these three measures. Figures 4.7 – 4.9 display the dispersion of scores for the ASD and control group on Coherence, Inappropriate Initiation, and Use of Context. One ASD child was identified as an outlier on Coherence and one was identified as an outlier on Use of Context. One control child was identified as an outlier on Inappropriate Initiation. The length of the ASD group’s box plot and the spread of the middle 50 % of the scores is less than the control group’s for Coherence and Use of Context indicating less score dispersal in the ASD group. The length of the box plots and the spread of the middle 50 % of scores for Inappropriate Initiation are similar for the two groups, indicating both groups have similar score dispersals on this measure.

91

14 12 10 8

13

6

Coherence 4 2 0 -2

N=

19

16

ASD

Control GROUP

Figure 4.7.


and control groups’ performance on Informants Number 1 rating for Coherence.

92

20

Inappropriate 10

Initiation 0

17

-10

N=

19

16

ASD

Control GROUP

Figure 4.8.


and control groups’ performance on Informants Number 1 rating for Inappropriate Initiation.

93

16 14 12

Use of

10

18 8

Context

6 4 2 0 -2

N=

19

16

ASD

Control GROUP

Figure 4.9.


and control groups’ performance on Informants Number 1 rating for Use of Context.

The standard deviations for the ASD group on Stereotyped Language and Nonverbal Communication are larger than the control group, suggesting more score dispersal in the ASD group. Figures 4.10 - 4.11 display the dispersion of scores for the two groups for these two measures. One ASD child was an outlier on Stereotyped Language. The length of the box plots for the two groups for Stereotyped Language is similar indicating similar score dispersal, and the length of the ASD group’s box plot for Nonverbal Communication is greater than the control group indicating greater dispersion of scores for the ASD group.

94

14 12

Stereotyped

10

13

Language

8 6 4 2 0 -2

N=

19

16

ASD

Control GROUP

Figure 4.10.


and control groups’ performance on Informants Number 1 rating for Stereotyped Language.

95

14 12 10

Nonverbal

8 6

Communication 4 2 0 -2

N=

19

16

ASD

Control GROUP

Figure 4.11.


and control groups’ performance on Informants Number 1 rating for Nonverbal Communication.

4.5.3

Clusters of Children with ASD Based on Pragmatic Measures

4.5.3.1 Clusters of children with ASD based on performance on the RHLB. Current status, based on the performance on the pragmatic tasks assessed by the RHLB, formed the basis for the potential clustering within the ASD group. All six RHLB measures were included for the cluster analysis. In order to obtain an optimal classification of ASD children, an agglomerative hierarchical cluster analysis was undertaken. The hierarchical clustering of the ASD child participants is shown in Figure 4.12.

96

Figure 4.12.

Agglomerative hierarchical cluster analysis of the 20 ASD

children based on performance on the six subtests from the Right Hemisphere Language Battery.

97 Four subgroups can be identified from the analysis. The resultant subgroups are descriptive only. Table 4.6 displays the membership of the subgroups. Scores are given as within standard deviations from the control group means. Performance scores within each subgroup support the partitioning decision.

98

Table 4.6

Subgroups of ASD children based on performance on the RHLB

Subgroup

ID

Given

Lge Hist

MP

WM CIM

AH

LS

PES

diag Subgroup 1 (n = 5)

9

AS

Normal

+1

+1

-1

-2

+1

-4

M age: 11, 7 (yrs, mths) (SD: 2, 5)

18

ASD

Delayed

0

+1

-1

-2

+1

-4

M TONI-2: 100 (SD: 15)

14

AS/ASD

Delayed

-1

+1

-2

-3

+1

-2

11

AS

Delayed

-2

0

-1

-1

+1

-2

20

ASD

Normal

-1

+1

-2

-1

+1

+1

Subgroup 2 (n = 3)

2

ASD

Normal

+1

+1

+1

+1

+1

-2

M age: 12, 3 (SD: 1, 5)

4

ASD

Normal

+1

+1

+2

+2

+1

+1

M TONI-2: 119 (SD: 14)

3

AS

Normal

+1

+1

+1

+2

-5

+1

Subgroup 3 (n = 5)

8

AS

Normal

+1

+1

+1

+1

+1

-5

M age: 13, 0 (SD: 2, 5)

13

AS

Delayed

+1

+1

-1

+2

+1

-5

M TONI-2: 106 (SD: 16)

6

AS

Normal

+1

+1

-1

0

+1

-3

15

AS

Normal

+1

+1

-2

+1

+1

-4

16

AS/ASD

Delayed

+1

+1

-2

+1

-5

-2

Table continues

Table 4.6 continued Subgroup

ID

Given

Lge Hist

MP

WM CIM

AH

LS

PES

diag Subgroup 4 (n = 7)

5

AS

Normal

+1

-2

-2

-1

-10

-1

M age: 10, 3 (SD: 1, 7)

7

AS

Delayed

-1

-2

-2

-2

-10

-3

M TONI-2: 87 (SD: 14)

19

ASD

Unsure

-3

-4

-2

-2

-10

+1

1

AS

Normal

-2

-2

-3

-1

-5

-4

12

AS

Unsure

-3

-3

-2

-2

+1

-3

10

AS

Delayed

-2

-2

-4

-2

+1

-7

17

ASD

Delayed

-2

-4

-4

-3

+1

-9

Note. ASD = Autism spectrum disorder. RHLB = Right Hemisphere Language Battery. AS = Asperger syndrome. Given diag = Independently determined diagnosis. Lge hist = Developmental language history. Normal = Language onset prior to age two years. Delayed = Language onset after two years of age. M P = Metaphor Picture Test. W M = Written Metaphor Test. C I M = Comprehension of Inferred Meaning. A H = Appreciation of Humour. L S = Lexical Semantic Test. P E S = Production of Emphatic Stress. TONI-2 = Test of Nonverbal Intelligence-2nd Edition. Yrs, mths = Years, months. Scores are given as standard deviations from the control means. + = Standard deviations above the control mean. - = Standard deviations below the control mean.

99

100

Subgroup 1 had generally above-average scores on Written Metaphor Test, mild difficulties on Metaphor Picture Test and Comprehension of Inferred Meaning, a range from mild to severe difficulties on Appreciation of Humour, above-average performance on Lexical Semantic Test, and generally moderate difficulties on Production of Emphatic Stress. Subgroup 2’s performance was generally average to above average on all six measures. Subgroup 3 had generally average skills on Metaphor Picture Test, Written Metaphor Test, Appreciation of Humour, and Lexical Semantic Test, mild to moderate difficulties with Comprehension of Inferred Meaning, and moderate-severe difficulties with Production of Emphatic Stress. Subgroup 4 had generally moderate-severe difficulties on all measures. Nonparametric tests (Kruskal-Wallis) established there was no significant difference between clusters on age (χ2 = 4.802; p = .187) or the TONI-2 (χ2 = 7.325; p = .062). 4.5.3.2 Clusters of children with ASD based on the CCC-2 ratings. An agglomerative hierarchical cluster analysis was undertaken using Informants Number 1 ratings on the five CCC-2 scales for the ASD children. The hierarchical clustering is shown in Figure 4.13. Two subgroups can be identified from the analysis. The resultant subgroups are descriptive only. Table 4.7 displays the membership of the subgroups. Scores are given as within standard deviations from the age-adjusted scaled score means for the control group on Informants Number 1 ratings on the five CCC-2 scales. The standard deviations above or below the control group means calculated for the individual ASD children support the partitioning decision. The bulk of the ASD participants (89 %) formed one subgroup. This subgroup was characterised by generally moderate to severe impairments across all five CCC-2 scales relative to the control group. The two outliers noted in the box plots in section 4.5.2.2 formed the second subgroup. Within this small subgroup, skill on all five CCC-2 scales approximated the scores of the control group.

101

Figure 4.13.


children based on Informants Number 1 ratings on the five scales from the Children’s Communication Checklist-Second Edition.

102

Table 4.7

Subgroups of ASD children based on Informants Number 1 ratings on the CCC-2

Subgroup

Subgroup 1 (n = 17)

ID

Coherence

Inappropriate Stereotyped

Use of

Nonverbal

Initiation

Language

Context

Communication

14

-2

-2

-3

-2.5

-3.5

20

-2

-2

-2.5

-2.5

-3

1

-2

-2.5

-3

-2.5

-3

12

-3

-2

-2.5

-2.5

-3

19

-2.5

-2.5

-1.5

-3

-2.5

7

-2.5

-3

-3.5

-3.5

-4

11

-3

-3

-4

-3.5

-4.5

4

-2

-2.5

-3

-3

-4

6

-2

-2.5

-4

-3

-4

15

-3

-3

-3.5

-2.5

-4.5

5

-3

-2

-3

-3.5

-4.5

10

-2.5

-1.5

-3.5

-3

-3.5 Table continues


Subgroup 1 continued

Subgroup 2 (n = 2)

ID

Coherence

Inappropriate Stereotyped

Use of

Nonverbal

Initiation

Language

Context

Communication

3

-2.5

-1.5

-3.5

-2

-4.5

8

-2

-1

-4.5

-2.5

-4

2

-3

-3

-5

-3.5

-5

16

-3.5

-3.5

-4.5

-3.5

-5

17

-3

-3

-5

-3.5

-3.5

13

-1

-2

0

-1

-2.5

18

-1.5

-1.5

-1.5

0

-1

Note. ASD = Autism spectrum disorder. CCC-2 = Children’s Communication Checklist-2nd Edition. Scores are given as standard deviations from the age-adjusted scaled score means for the control group. + = Standard deviations above the control mean. - = Standard deviations below the control means.

103

104 4.6

Discussion

Chapters 2 and 3 of this thesis describe linguistic deficits and difficulties with language skills that emerge subsequent to the initial onset of language, reflecting Bloom and Lahey’s (1978) language content and form, in the group of children with ASD investigated in the series of studies. The current study aimed to provide clinically useful descriptions of their use of language by investigating verbal and nonverbal pragmatic skills. 4.6.1

Pragmatic Skills in Children with ASD Reclassified as AS or HFA According to Reported Developmental Language History

Following the methodological approach used in previous validation studies (e.g., Howlin, 2003; Mayes & Calhoun, 2001), the current study applied DSM-IV’s (APA, 1994) language onset criterion to reclassify the ASD participants as AS or HFA. Subjective ratings from the CCC-2 (Bishop, 2003) scale relating to pragmatic language (e.g., Use of Context) indicated that AS and HFA cannot be differentiated on pragmatic language skills. CCC-2 questions assessing the use of language in context cover humour appreciation, literal interpretation of language, and the understanding of irony. Based on the perception of others familiar with the children, the reclassified AS and HFA groups were similarly proficient on a range of other pragmatic functions such as initiation and coherence in discourse, the use of stereotypical language, and nonverbal skills such as the use and interpretation of facial expression and eye contact, and the use of gestures. The findings from the investigation of proficiency on objective pragmatic language tasks generally support the subjective findings. Based on performance on the RHLB (Bryan, 1989), the findings suggest that children reclassified as AS in the current study could be differentiated from the children reclassified as HFA on one pragmatic language measure only. The RHLB includes a number of pragmatic language tasks, such as metaphorical language, inferential language, and appreciation of humour, but only the interpretation of pictorially presented metaphors differentiated the reclassified groups of ASD children. The present findings offer support for previous research findings. The lack of

105 significant difference between AS and HFA on any of the CCC-2 scales (Bishop, 2003) included in the present study is consistent with findings by Verte, Geurts, Roeyers, Rosseel, et al. (2006), who used an earlier version of the CCC-2 in their validation study. Moreover, the production of variations in emphatic stress to convey contrastive meaning in speech, as assessed by the RHLB (1989), failed to differentiate the two presentations. This finding is consistent with Ramberg et al.’s (1996) results, where no distinction could be made between the two disorders on intonation patterns in speech. Nonetheless, the lack of significant findings between the AS and HFA groups on aspects of discourse, such as coherence, inappropriate initiation, and the use of stereotypical language as determined by the subjective ratings from the CCC-2 (Bishop, 2003), is inconsistent with findings from previous studies (e.g., Fine et al., 1994; Gilchrist et al., 2001; Shriberg et al., 2001). It may be that, in the current study, the informants familiar with the children were accustomed to their conversational style and overlooked possibly deviant discourse behaviour. Objective analysis is needed to clarify the discourse behaviours in the reclassified AS and HFA groups of children. Previous studies (Howlin, 2003; Mayes & Calhoun, 2001) have offered consistent results suggesting developmental language history is not a significant factor in subsequent linguistic and language outcome in ASD. However, the results of studies (including the current study) where the classification of participants into AS and HFA has been based on developmental language history should be interpreted cautiously, as discussed previously (refer to sections 2.6.1 and 3.6.1). 4.6.2 Pragmatic Skills in Children with ASD The children, combined as a non-delineated ASD group, presented with highly significant differences in pragmatic skills relative to the control group on all CCC-2 scales (Bishop, 2003) included in the present study. The differentiation of the ASD group from a group of typically developing children supports the use of the CCC-2 as a screening measure of pragmatic language. The combined ASD group also experienced significant difficulties with some aspects of the RHLB (Bryan, 1989)

106 compared to the typically developing control group. The ASD group of children presented with reduced skills in comprehension of inferential language, production of emphatic stress to convey contrastive meaning in speech, and linguistic flexibility, as assessed by the Lexical Semantic Test. The identification of difficulties with understanding inference and ineffective use of stress is consistent with previous findings (Dennis et al., 2001; Paul et al., 2005; Shriberg et al., 2001). The present study and the study undertaken by Gunter et al. (2002) both utilised a number of the same subtests from the RHLB (Bryan, 1989). Nevertheless, there are inconsistencies between the two studies’ findings. The Metaphor Picture Test, the Written Metaphor Test, Comprehension of Inferred Meaning, and Appreciation of Humour were subtests common to both studies. While the current study identified the non-delineated ASD group as possessing intact metaphorical language skills and humour appreciation, but deficits in comprehending inferential language, Gunter et al. identified intact metaphorical and inferential language skills, but deficits in humour appreciation. The inconsistencies between the findings of the current study and Gunter et al.’s (2002) findings may be due to methodological differences. Gunter et al.’s study consisted of seven children and one adult, all of whom met the ICD-10 (WHO, 1992) criteria for AS, whereas in the present study, children with AS and HFA were included. The debate regarding the categorical distinction between AS and HFA remains unresolved (see e.g., Klin et al., 2005; Verte, Geurts, Roeyers, Rosseel, et al., 2006). Findings from this thesis thus far suggest that AS and HFA can be differentiated on two aspects of language (resolving ambiguity and interpreting pictorially presented metaphorical language) (refer to sections 3.6.1 and 4.6.1 respectively). It is possible, therefore, that the differing diagnoses used by Gunter et al. and the present study could account for the discrepant findings. Performance on some of the objectively assessed pragmatic language skills failed to differentiate the ASD group from the control group. As assessed by the RHLB (Bryan 1989), the ASD group was just as competent as the control group with interpreting metaphorical language and appreciating humour. Nonetheless, the subjective ratings from the CCC-2 (Bishop 2003) identified pragmatic language (Use

107 of Context scale) as one aspect of communication that differentiated the ASD group from the control group. Although previously used by Gunter et al. (2002) with a paediatric group, the RHLB is designed for the adult population. Therefore, it may be that the pragmatic language tasks of interpreting metaphorical language and humour appreciation were equally hard for both groups of children. The theoretical perspective of combining AS and HFA into a single experimental group adopted in the present study failed to acknowledge the importance of examining within-group differences in the ASD population. Tager-Flusberg (2004), for instance, has argued that important clinical differences may be obscured in such research approaches. The pragmatic skills subjectively profiled in the CCC-2 (Bishop, 2003) scales used in the current study are central characteristics of the autistic presentation (APA, 2000). It is, therefore, not surprising that the ASD group presented with similar or more homogeneity than the control group on four of the five rating scales. In contrast, when specific verbal and nonverbal pragmatic skills were objectively investigated using the RHLB (Bryan, 1989), greater heterogeneity was noted in the ASD group relative to the control group. Acknowledging, but not further investigating, the range of specific pragmatic deficits in a combined AS/HFA study may provide research findings with limited clinical relevance only. 4.6.3 Heterogeneity of Pragmatic Skills in Children with ASD Despite the conceptualisation of a spectrum or continuum of autistic disorders, few studies (Prior et al., 1998; Verte, Geurts, Roeyers, Rosseel, et al., 2006) have investigated the heterogeneous nature of communication skills within ASD. Subgroups, based on linguistic and metalinguistic measures, have been described in the ASD children investigated in the series of studies presented thus far in this thesis (see sections 2.6.3 and 3.6.3). In the current study, investigations of subgroups within the ASD group of children were undertaken using broad subjective measures and specific objective measures of pragmatic skills. Two subgroups within the ASD group of children could be described using the CCC-2 measures (Bishop, 2003). The description of one large homogeneous group with subjectively impaired pragmatic skills accords with the diagnostic requirement of

108 qualitative impairment in social interaction for both AD and AS (APA, 2000). Of interest, however, is the second subgroup identified within the ASD group of children. This subgroup, consisting of two participants, was characterised as having CCC-2 ratings that were close approximations of the control group’s means. Previous analyses identified that the two ASD participants forming the second subgroup have relatively intact language skills. Participant 13, in particular, possesses language skills that are commensurate with his typically developing peers (refer to sections 2.5.3 and 3.5.3). A possible explanation for their distinct pragmatic profile may be that their competence with language influences how others perceive them in social interactions. This, however, could not be assessed in the current study. Alternatively, it is possible that the two children in the second subgroup are not, in fact, ASD. Participant 13 was diagnosed AS by a psychiatrist, and participant 18 was diagnosed ASD by a paediatrician (see Table 2.1). Both participants have a history of delayed onset of language. The possibility that diagnosticians may be disregarding developmental language history when making diagnostic decisions regarding autistic classification was raised in section 2.6.1. It is possible that other DSM-IV (APA 1994) diagnostic criteria, such as the criterion relating to social interaction, may not be adhered to strictly as well. Rouse (2004), for instance, reported on a survey of psychiatrists and paediatricians in Queensland, Australia. Fifty-eight percent of the 105 respondents reported having upgraded a child’s symptoms to meet ascertainment categories required for educational assistance. In contrast to the large, overwhelmingly homogeneous subgroup resulting from the cluster analysis using the five CCC-2 scales (Bishop, 2003), a number of subgroups within the ASD group could be described based on the objective pragmatic measures examined by the RHLB (Bryan, 1989). Nonverbal cognition was a significant factor in the formation of both the linguistic (refer to section 2.6.3) and metalinguistic (see section 3.6.3) subgroups of children with ASD. In the present study, however, nonverbal cognitive skill was not a factor in determining the formation of the subgroups. The inconsistent findings of nonverbal cognitive influences over the heterogeneous language skills in children with ASD warrants further investigation.

109 While heterogeneity of pragmatic proficiency was noted within the ASD group, a range of competencies within the individual was also evident, suggesting that individuals with ASD may present with differing strengths and weaknesses on pragmatic tasks. ASD participant 15, for instance, scored one standard deviation above the control group mean on interpreting metaphorical language, humour appreciation, and lexical-semantic flexibility, but two standard deviations below on understanding inferential language and four standard deviations below on producing prosodic variations in speech. Participant 9’s strengths were in interpreting metaphorical language and lexical-semantic flexibility (one standard deviation above the control means) while interpreting inferential language, appreciating humour, and producing emphatic stress were weak pragmatic skills (between one to four standard deviations below the control means). Participant 7’s performance on interpreting written metaphorical language was a strength but performance on all other RHLB (Bryan, 1989) tasks suggested areas of weakness. A small number of ASD participants presented with average to above skills on the RHLB (Bryan, 1989) in the present study (subgroup 2). These participants (ASD 2, 3, and 4) were clustered into the most competent subgroup in the present study, and clustered together into the most competent subgroups on linguistic (see section 2.5.3) and metalinguistic (see section 3.5.3) tasks, suggesting well developed competencies across a range of communication parameters. In contrast, ASD participant 8 was identified in sections 2.5.3 and 3.5.3 as experiencing generally moderate language difficulties on the CELF-4 (Semel et al., 2003) and the TLC-E (Wiig & Secord, 1989). Despite these difficulties, his pragmatic language skills are above average. Undertaking comprehensive language and communication assessments coupled with an examination of the heterogeneity of skills within ASD may provide a greater clinical appreciation of the strengths and weaknesses within the autism spectrum, and also within the individual child. Erroneous clinical assumptions that may lead to ineffectual or unnecessary interventions are avoided if an individualised approach to assessment is adopted for children with ASD.

110 4.7

Specific Limitations of the Study

Although the RHLB (Bryan, 1989) was designed for the neurologically-impaired adult population, it has been used in research with children with autistic symptomatology (see Gunter et al., 2002). The RHLB administration manual includes a table to convert adult raw scores to relative T Scores. For the present study, however, the children’s raw scores were used for the analyses. These scores were not corrected for age. The findings may, therefore, reflect age differences rather than differences in actual pragmatic language skills. Further, the RHLB has been rated poorly in terms of content validity, construct validity, and reliability (Tompkins, 1995). The limited data on the psychometric properties of the RHLB may undermine confidence in the results of the study. Nevertheless, no other instrument exists that investigates as many different pragmatic behaviours as does the RHLB. Therefore, its use provides some measure of identifying variability of performance between and within groups. Additionally, the present study provides no adequate explanation for the identification of the anomalous group of two ASD children classified according to subjective ratings of pragmatic skills. The use of a validation instrument such as the ADOS-G (Lord et al., 2000) may clarify the validity of the two children’s diagnoses. There are a number of nonverbal and pragmatic communication skills not assessed in the current analysis. Compromised nonverbal skills, such as eye contact, have been described in the ASD population (Njiokiktjien et al., 2001; Scott, 1985; Tantam et al., 1993). An assessment of these skills would complement the present study. Impaired discourse skills have likewise been described in ASD (e.g., Adams, Green, Gilchrist, & Cox, 2002; Gilchrist et al., 2001). The RHLB Discourse Analysis (Bryan, 1989), not included in the present study, evaluates a range of skills in a twoway conversation, and covers aspects of discourse such as the use of supportive routines (e.g., greetings), questions to gain information, and assertive routines (e.g., changing the behaviour of the conversational partner by making complaints, demands, criticisms, and giving advice), and an evaluation of the length of a speaker’s turn and the level of detail provided, topic content, types of interactions (e.g., providing information, expressing opinions), level of formality, turn taking abilities, and meshing. The present study would benefit from the inclusion of an assessment of discourse skills, particularly as discourse skills may differentiate AS from HFA (see

111 e.g., Fine et al., 1994; Ghaziuddin & Gerstein, 1996; Gilchrist et al., 2001). Future research should aim to profile these nonverbal and pragmatic skills using video analysis of conversational and interaction skills. 4.8

Conclusion

The current study applied the three theoretical research positions presented in section 1.3 to investigate a range of pragmatic skills in children with ASD. Objective data on pragmatic skills were obtained using the RHLB (Bryan, 1989). Subjective data were obtained from adults familiar with the child using the CCC-2 (Bishop, 2003). Performance on one RHLB measure only and no measures from the CCC-2 differentiated AS from HFA, when the children were reclassified as AS or HFA using DSM-IV’s (APA, 1994) language criterion. When regrouped into a diagnostically undifferentiated ASD group, a number of the RHLB measures failed to differentiate the ASD group from the control group. The ASD group, however, was identified as having compromised skills on the interpretation of inferential language, linguistic flexibility, and the production of emphatic stress as assessed by the RHLB relative to the control group. All CCC-2 measures differentiated the two groups. Greater heterogeneity within the combined ASD group than in the control group was evident on all RHLB tasks, and on the Nonverbal Communication scale of the CCC-2. Cluster analysis, based on the subjective CCC-2 (Bishop, 2003) ratings, revealed one large, generally homogeneous subgroup, and one small subgroup comprised of two outliers within the ASD group of children. In contrast, cluster analysis, based on performance on the objective RHLB (Bryan, 1989) pragmatic tasks, identified a number of subgroups within the ASD group of children. Proficiency within the subgroups ranged from average on all measures to severe difficulties on all measures. The findings of a range of proficiencies within the ASD group and within the individual on pragmatic skills as assessed by the RHLB suggest that intervention and support would need to be individually tailored to a child’s profile. Individual assessments would, therefore, be necessary to optimise the child’s social and academic development. Hence, clinically relevant data on pragmatic skills in ASD were not revealed until within-group analyses, reflecting the conceptualisation of ASD as initially proposed by Wing (1989), were undertaken.

112 Chapters 2 – 4 of the present thesis examine fundamental, abstract, and pragmatic language skills in children with ASD. Manjiviona (2003), however, has suggested that an examination of verbal problem solving skills should likewise be included in communication assessments undertaken with this population. To further extend the clinical understanding of language and communication skills associated with AS and HFA, the findings from an assessment of social problem solving skills are reported in the following chapter.

113

5

Social Problem Solving in Children with Autism Spectrum Disorder

5.1

Introduction

Lord (2000) has argued that deficient language skills may contribute to difficulties in areas of academic, behavioural, and social development in children with a diagnosis of ASD. Other researchers (e.g., Solomon, Goodlin-Jones, & Anders, 2004) have proposed that deficiencies in other skill areas may likewise contribute to reduced developmental outcomes. Solomon et al., for instance, have suggested that deficient social problem solving skills may contribute to the social-emotional difficulties observed in children with autistic symptomatology. It is, therefore, not surprising that Manjiviona (2003) recommended that assessments of communication skills associated with AS, HFA, or ASD include an examination of social communication skills such as verbal or social problem solving. For professionals charged with providing intervention to children with AS, HFA, or ASD, such as speech-language pathologists, an accurate description of problem solving in everyday life situations would be of clinical relevance (American Speech-Language-Hearing Association, 2000). 5.2

Studies Investigating Social Problem Solving in Children with ASD

As reflected in chapters 2 – 4 of this thesis, a large number and range of studies, encompassing the theoretical positions of the validation, combined, and spectrum approaches, have been undertaken to define language and communication skills in children with AS, HFA, or ASD. In contrast, few studies have focused on social problem solving skills in these populations. There are two known studies (Channon et al., 2001; Griswold et al., 2002), and both investigated social problem solving in diagnostically homogeneous groups of children.

114 5.2.1

AS-specific Studies

Both known studies (Channon et al., 2001; Griswold et al., 2002) focused on children with diagnoses of AS as defined by DSM-IV criteria (APA, 1994). Participants in Channon et al.’s study were aged between 11 – 19 years (mean age: 13.89 years) and all had nonverbal IQs in the normal range (70 or above). The AS children were matched to typically developing children on age, nonverbal mental ability, and receptive and expressive language scores. A short version of the Predicaments problem solving task (Channon & Crawford, 1999) was used to assess real-life problem solving skills. The Predicaments test provides information on the ability to provide a factual account of the problem situation, generate a number of solutions, select optimal and personally preferred solutions, rate the degree of satisfaction with the optimal and personal solutions, and rank the alternative solutions from best solution to worst solution. Additionally, responses are scored in terms of problem appreciation, the social appropriateness of the solution, the practical effectiveness of the solution, and the number of alternative solutions generated. The AS group of participants in Channon et al.’s (2001) study was characterised by difficulties in generating high quality solutions to solve the problems, and their optimal and personally preferred solutions were of a poorer quality, particularly in terms of social appropriateness of the solutions, relative to the control group. These findings appear to support the inclusion of a focus on social problem solving behaviours, such as problem identification and prioritisation, generation of feasible resolution strategies, and evaluation of individual strategies, in social adjustment enhancement intervention programmes for children with ASD (see e.g., Solomon et al., 2004). The focus of Griswold et al.’s (2002) study was social problem solving and academic achievement in children and adolescents with AS. It is, therefore, difficult to make comparisons between Griswold et al.’s findings and those of Channon et al. (2001). Griswold and colleagues administered the Test of Problem SolvingElementary, Revised (Zachman, Huisingh, Barrett, Orman, & LoGiudice, 1994) or the Test of Problem Solving-Adolescent (Bowers, Barrett, Huisingh, Orman, & LoGiudice, 1991), and the Wechsler Individual Achievement Test (WIAT)

115 (Psychological Corporation, 1991) to children and adolescents aged between 6 years 9 months – 16 years 8 months (mean age: 10, 0 months). Griswold et al. (2002) described a significant mismatch between performance on the WIAT (Psychological Corporation, 1991) and the tests of problem solving for the children. Further, Griswold and colleagues noted considerable variability in performance scores within the diagnostically homogeneous group of children on the tests of problem solving. The mean Standard Score was 73.52, but scores ranged from 55 – 118. Griswold et al.’s (2004) results offer clinicians insight into the degree of difficulty children with AS may experience in social problem solving situations. The findings, however, should be viewed as preliminary findings only, as there are a number of limitations to the study that require redressing. Firstly, Griswold and coresearchers stated they wanted to determine the relationship between verbal comprehension and expression and problem solving skills. Comparisons, rather than correlations, however, were undertaken by the research group. Griswold et al.’s interest lay in the discrepancy between language skill and problem solving skill, which may explain why correlations were not investigated. An examination of the relationship between language skill and problem solving, however, would be highly relevant as previous researchers have argued that language plays an influential role in critical thinking, reasoning, and problem solving (Baldo et al., 2005; Liss et al., 2001; Solomon et al., 2004; but see also Joseph, McGrath, & Tager-Flusberg, 2005). Secondly, Griswold et al.’s (2002) finding of diverse social problem solving skills within a diagnostically homogeneous group warrants further investigation. Although all participants were independently classified by psychiatrists, psychologists, or medical practitioners as AS using DSM-IV criteria (APA 1994), numerous concerns have been raised regarding the accuracy and reliability of such diagnoses (see section 2.6.1). Klin et al. (2005), for instance, have cautioned that the DSM-IV distinction between AS and HFA may not consistently be applied in diagnostic and clinical practice. As discussed in section 2.6.1 of this thesis, there was a lack of strict adherence to DSM-IV diagnostic criteria by psychiatrists, paediatricians, and psychologists noted regarding the diagnoses of the children

116 investigated in the present study. The heterogeneity noted by Griswold et al. (2002) may, therefore, be due to the inclusion of other pervasive developmental disorders. It is possible that Griswold et al.’s (2002) study included participants with HFA. If HFA and AS have similar difficulties with social problem solving, the inclusion of both diagnoses may not necessarily affect the results. If, on the other hand, AS and HFA present with different profiles in social problem solving skills, this could account for the diverse skills documented by Griswold et al. No study to date has applied the framework of the validation approach to determine whether AS and HFA can be differentiated on social problem solving abilities. 5.3


Channon et al.’s (2001) findings of specific problem solving deficits appear to provide clinically relevant data that can be used for the development of intervention programmes. The theoretical approach of artificially segregating aspects of problem solving and critical thinking, as used by Channon and colleagues, however, has been challenged by some research groups (Bowers et al., 1991; Zachman et al., 1994), who argued that the skills required for problem solving are diverse, complicated, and highly interrelated. In contrast to the study undertaken by Channon et al., the problem solving assessments chosen by Griswold et al. (2002) reflect the view that the skills required for social problem solving are interwoven. While Griswold et al. (2002) described reduced skills in social problem solving in children with AS, the utility of their findings to clinical practice may be limited. Language skill may influence social problem solving (Baldo et al., 2005; Liss et al., 2001; Solomon et al., 2004), but no examination of the relationship between language skill and social problem solving skill was undertaken by Griswold and coresearchers. Further analysis is required to clarify the relationship between language and social problem solving. Heterogeneity of social problem solving skills was noted by Griswold et al. (2002) within a diagnostically homogenous (AS) group of children. Mawhood, Howlin, and Rutter (2000) argued that examining the cause of the heterogeneity within ASD may be a means of gaining insight into the disorder. It may be that,

117 despite the DSM-IV (APA, 1994) diagnostic classification of participants as AS, Griswold et al.’s group of children included other pervasive developmental disorders such as HFA. It has not yet been established if AS and HFA are similar expressions of autism (see e.g., Frith, 2003; Macintosh & Dissanayake, 2004). If tasks of social problem solving differentiate AS from HFA, this could account for heterogeneity within an (inadvertently) combined experimental group. 5.4


The overall aim of the present study was to extend Griswold et al.’s (2002) findings. Firstly, the ASD children were reclassified as AS or HFA based on reported developmental language history (as per the framework from the validation studies outlined in section 1.3.1) to determine whether performance on tasks of social problem solving could differentiate AS from HFA. Secondly, the theoretical approach of the combined studies, as outlined in section 1.3.2, was applied. The Test of Problem Solving-Elementary, Revised (TOPS-E, R) (Zachman et al., 1994) or the Test of Problem Solving-Adolescent (TOPS-A) (Bowers et al., 1991), the two tests used by Griswold et al., were administered to an ASD group of children and a group of typically developing control children. To extend Griswold et al.’s study, the correlations between language skill and social problem solving for both groups of children were investigated. The effect of language skill was removed to allow for a comparison of social problem solving skills in the ASD group relative to control group. Score dispersal within the ASD and the control group was investigated to examine heterogeneity. Although the previous studies reported in this thesis included cluster analyses to examine within-group differences, this theoretical approach was not undertaken in the present study as cluster analyses cannot be undertaken with one measure only (in this case the measure being performance on TOPS-E R /TOPS-A). Due to an absence of validation studies investigating the social problem solving skills of AS and HFA, no prediction was made regarding a distinction between the two reclassified groups of children (i.e., AS and HFA) based on performance on social problem solving tasks. Based on the suggestion that language skill may influence social problem solving skill (Baldo et al., 2005; Liss et al., 2001), it was predicted that there would be significant correlations between language skill

118 and performance on the TOPS-E R (Zachman et al., 1994) or TOPS-A (Bowers et al., 1991) for the ASD group and the control group. Additionally, it was predicted that performance by the ASD group on the tests of social problem solving would differ significantly from the control group when the effect of language skill was removed. No prediction was made regarding heterogeneity of skill on social problem solving within the ASD group. 5.5

Method

5.5.1

Participants

The 20 children with average intelligence and diagnoses of AS, HFA, and/or ASD and the 18 typically developing control children described in section 2.4.1 were involved in the present study (demographic details can be found in Table 2.1). 5.5.2

Measures

All children were assessed with the TOPS-E, R (Zachman et al., 1994) or the TOPS-A (Bowers et al., 1991), the TONI-2 (Brown et al., 1990), and the CELF-4 (Semel et al., 2003). The TONI-2 and the CELF-4 are described in an earlier section of this thesis (refer to section 2.4.2). The TOPS-E, R is a diagnostic test of problem solving designed to assess language-based critical thinking skills (Zachman et al., 1994). The test is suitable for children aged 6 years 0 months – 11 years 11 months. The psychometric properties of the TOPS-E, R are satisfactory. According to Zachman et al., the internal consistency of the test, based on homogeneity of items, yielded an average reliability coefficient across the age groups of .85, test-retest reliability of .88, and the average validity for all items across all ages established that 81 % of the items resulted in statistically significant average correlations with total test scores. The TOPS-E, R consists of 14 pictured scenarios and 72 questions. An example of one scenario and the related questions are presented in Appendix G. The TOPS-A is a diagnostic test of problem solving designed to assess language-based critical thinking skills in children aged 12 years 0 months – 17 years 11 months (Bowers et al., 1991). The psychometric properties of the TOPS-A are

119 adequate. Bowers and colleagues advise the internal consistency of the test, based on homogeneity of items, yielded an average reliability coefficient across the age groups of .85, test-retest reliability of .86, and the average validity for all items across all ages established that 82 % of the items resulted in statistically significant average correlations with total test scores. The TOPS-A consists of 13 scripts and 50 questions. The examiner reads the script to the participant while the participant simultaneously reads the script. The examiner then asks questions relating to the script. The script and the written questions remain in view of the participant. An example of a script and the related questions for the TOPS-A are shown in Appendix H. 5.5.3 Procedure The TOPS-E, R (Zachman et al., 1994) and the TOPS-A (Bowers et al., 1991) manuals provide strict guidelines regarding what must be alluded to in the response for the response to be acceptable. Examples of the TOPS-E, R and TOPS-A scoring guidelines are displayed in Appendix G and H respectively. Scoring judgments were made in strict accordance with the guidelines provided in the tests’ manuals. The children’s raw scores were converted to Standard Scores as per the instruction manuals. Due to age restrictions of the tests of problem solving, 12 of the children with ASD (age range 9 years 0 months – 11 years 11 months) and 11 of the control children (age range: 9 years 1 month – 11 years 10 months) were assessed with the TOPS-E, R (Zachman et al., 1994). The remaining eight children with ASD (age range: 12 years 1 month – 17 years 1 month) and the remaining seven control children (age range: 12 years 7 months – 14 years 10 months) were assessed using the TOPS-A (Bowers et al., 1991). The Standard Scores of both problem solving assessments reflect a standard measure. Therefore, scores from the TOPS-E, R and the TOPS-A were entered as TOPS (Tests of Problem Solving) for the statistical analyses. Unless otherwise stated, both tests are referred to as TOPS for the remainder of the study. The reclassification procedure outlined in section 2.4.3 was applied for the first aspect of the present study in order to examine differences in social problem

120 solving performance between AS and HFA. The AS group was similar to the HFA group on age (Z = -1.022; p = .307), core language skills (Z = -1.200; p = .230) as assessed by the Core Language Index from the CELF-4 (Semel et al., 2003), and nonverbal cognitive skills (Z = -0.867; p = .386) as assessed by the TONI-2 (Brown et al., 1990) (see section 2.5.1). Due to the small sample size, statistical comparison of the TOPS (Bowers et al., 1991; Zachman et al., 1994) means only was undertaken for the reclassified AS and HFA groups. The undifferentiated ASD group of participants investigated in the present study had a similar mean age (t = 0.223; p = .825), significantly less proficient core language skills (t = -3.800; p = .001) as assessed by the Core Language Index from the CELF-4 (Semel et al., 2003), and significantly less proficient nonverbal cognitive skills (t = -2.259; p = .03) as assessed by the TONI-2 (Brown et al., 1990) relative to the control group (see section 2.5.2). Correlations between verbal skill and problem solving skills were determined for the combined study (undifferentiated ASD group). An analysis of covariance (ANCOVA), using the Core Language Index as covariate and the TOPS as the dependent variable, was undertaken to remove the effects of verbal skills. 5.6

Results

5.6.1

Performance on the TOPS by the Children with ASD Reclassified as AS or HFA Based on Reported Developmental Language History

Group means and standard deviations from the TOPS for the reclassified AS and HFA groups are shown in Table 5.1. For determination of statistical significance, a conservative level of p ≤ .05 was utilised. A non-parametric Mann-Whitney U test indicated that there was no significant difference between the two groups on their mean TOPS scores (p ≥ .05).

121

Table 5.1

Descriptive statistics of the performance on the TOPS

by children with ASD reclassified AS and HFA according to developmental language history Variable

Language

Language

onset prior to

onset after two Test

age two years

years

(AS) (n = 10)

(HFA) (n = 8)

M

M

SD

SD

Mann-Whitney U

Z

p (Asymp. Sig. 2 tailed)

TOPS

71

9

62

11

-1.949

.051

Note. ASD = Autism spectrum disorder. AS = Asperger syndrome. HFA = High functioning autism. TOPS = Test of Problem SolvingElementary, Revised or Test of Problem Solving-Adolescent.

5.6.2

Performance by all Child Participants on the TOPS

Table 5.2 displays the means, standard deviations, and ranges for the non-delineated ASD group and the control group for the Core Language Index and the TOPS. Pearson correlation coefficients were significant between the Core Language Index from the CELF-4 and performance on the TOPS for the ASD group (.646; p = .002) and the control group (.683; p = .002). ANCOVA identified a highly significant group difference (F (1, 36) = 17.215, p = .001) on the TOPS when the effect of language skill was removed. Of the two groups, the ASD group of children was significantly less skilled on social problem solving as assessed by the TOPS.

122

Table 5.2 Means, standard deviations, and ranges of scores for the child participant groups on Core Language (CELF-4) and the TOPS ASD

Control

(N =20)

(N =18)

Variable

M

SD

Range

M

SD

Range

Mean diff.

Core Lge

84

20

46 - 120

106

12

82 - 121

22

TOPS

66

11

< 55 - 83

90

13

55 - 105

24

Note. ASD = Autism spectrum disorder. Core Lge (CELF-4) = Core Language Index from the Clinical Evaluation of Language Fundamentals-4th Edition. TOPS = Test of Problem Solving-Elementary, Revised or Test of Problem Solving-Adolescent. Mean diff = Differences in the means of the two groups.

As shown in Table 5.2, the ASD group’s standard deviation on the TOPS is slightly less than that of the control group, suggesting the possibility that the ASD group’s performance scores are more concentrated. Figure 5.1 displays the spread of scores for the two groups. Descriptively, the box plot for the ASD group is shorter than the box plot for the control children, indicating less score dispersal in the ASD group. Fifty percent of the ASD scores are concentrated between Standard Scores of 75-85. In contrast, beyond the middle 50 % of the control group’s scores, the scores are spread between a Standard Score of 55 through to a Standard Score of 105.

123

110

100

TOPS

90

80

70

60

50

N=

20

18

ASD

Control GROUP

Figure 5.1.

Box plots displaying the dispersal of scores for the child ASD

and control groups’ performance on the tests of problem solving (Test of Problem Solving-Elementary, Revised or the Test of Problem SolvingAdolescent).

5.7

Discussion

The present study aimed to extend the findings from Griswold et al.’s (2002) study of social problem solving in children with AS. When the ASD children were reclassified as AS or HFA according to DSM-IV’s (APA, 1994) language criterion, performance on social problem solving tasks failed to differentiate between the two reclassified groups. The non-differentiated ASD group of children involved in the current study presented with poorer social problem solving skills when compared to the typically developing control group when the effect of language was controlled. 5.7.1

Social Problem Solving in Children with ASD Reclassified as AS or HFA According to Reported Developmental Language History

No known study has investigated whether children with a diagnosis of AS can be differentiated from children with a diagnosis of HFA based on performance on social problem solving tasks. Findings from the present study suggest performance on

124 problem solving tasks based on familiar social encounters does not distinguish AS from HFA. The lack of differentiation between the two disorders on social problem solving is consistent with the findings presented in chapters 2 - 4 of this thesis, where a range of language and communication tasks failed to differentiate AS from HFA on all but two measures. The tentative findings, therefore, suggest that even if participants with a diagnosis of HFA were included in Griswold et al.’s (2002) study investigating social problem solving in children with AS, the diversity of problem solving capabilities observed in the study cannot be attributed to their inclusion. Nevertheless, it is noteworthy that performance differences between the two reclassified groups of children approached significance in the present study. It is also noteworthy that five children reclassified as HFA, but only one reclassified as AS, were allocated the basal score of “less than 55”. It is possible that, with larger participant groups and Standard Scores that adequately reflect the performance of lower-scoring participants, differentiation between AS and HFA on social problem solving might be evident. As discussed in section 2.6.1, however, the current findings may reflect methodological problems associated with the reliance upon retrospective recall of developmental milestones for diagnostic purposes (Klin et al., 2005; Tager-Flusberg, 2003; Woodbury-Smith et al., 2005). Consequently, the use of parent-reported history as a diagnostic marker in validation studies may not provide accurate differentiation of children as being either AS or HFA. Thus, any problem solving behaviour identified as being related to either AS or HFA may be based on flawed diagnostic categories. 5.7.2

Social Problem Solving in Children with ASD

The group of children with AS or HFA, when viewed as a non-differentiated ASD participant group, presented with problem solving difficulties not present in the control children. This finding is consistent with Griswold et al.’s (2002) results which were based on the same diagnostic tests used in the current study. The findings, however, extend Griswold et al.’s work, as the influence of language skill over social problem solving was examined in the current study. For both the ASD group and the

125 control group, there was a highly significant correlation between core language skill and performance on the TOPS. The identification of a correlation between language skill and verbal problem solving supports recent suggestions that language skill may influence verbal problem solving (Baldo et al., 2005; Liss et al., 2001). Results from an assessment reported earlier in this thesis suggest that the ASD group’s core language skills were significantly impaired compared to the control group (see section 2.5.2). This, coupled with the highly significant correlation between language skill and performance on the TOPS, necessitated the controlling of language skill in the analysis. When language was controlled, the ASD group of children performed significantly below the control group on tasks involving social problem solving. Contrary to Griswold et al.’s (2002) findings, and the findings from the preceding chapters of this thesis where non-homogeneity of skill within the combined AS/HFA studies was evident, the results of the present study suggest greater homogeneity in overall problem solving skills within the non-delineated ASD group compared to the control group. This finding, however, is misleading. The lowest Standard Score for the TOPS is a score designated “less than 55”. Only one of the 18 control children obtained a Standard Score of “less than 55”. In contrast, eight of the 20 ASD children obtained a similar Standard Score. Hence, the box plots, based on Standard Scores, do not accurately reveal the spread of scores below the basal score for the tests. To offer clinical utility in interpreting the range of scores within a sample population on the TOPS, it would be necessary for specific Standard Scores to be allocated to account for children whose raw scores do not convert to a Standard Score of 55, such as the nine in the present study. 5.7.3 Clinical Application of the Findings The effect of verbal skill was controlled in the present study, as it was in Channon et al.’s (2001) study, and in both instances, the ASD/AS group of children presented with impaired social or verbal problem solving skills. It is, therefore, tempting to argue that the impaired performance by the children in both studies must be due to

126 deficits in problem solving alone, and that targeting problem solving in intervention programmes would be clinically justifiable. Other variables not accounted for in the present study, however, may have influenced the results. Everyday problem solving, for instance, draws on knowledge derived from direct experience and social learning (Channon, 2004), and children with ASD are commonly excluded from their peer culture (Wolfberg & Schuler, 2006). Impoverished life experiences may, therefore, have influenced the ASD group’s performance. Executive function skills may influence the success with which problem solving situations are resolved (Charman, 2004). As numerous studies have highlighted executive skill deficits in children with AS, HFA, and/or ASD (e.g., Goldberg et al., 2005; Hill, 2004; Joseph et al., 2005; Verte, Geurts, Roeyers, Oosterlaan, & Sergeant, 2006; Zelazo, Jacques, Burack, & Frye, 2002), executive function deficits may account for the poor performance by the children with ASD on the social problem solving tasks. In addition, problem solving draws on a range of cognitive processes (Channon et al., 2001). Everyday problem solving may involve applying an understanding of the situation from another person’s perspective (i.e., theory of mind) (Channon, 2004; Pennington & Ozonoff, 1996), and theory of mind (ToM) deficits have been described in the ASD population (e.g., Baron-Cohen et al., 1997). The problem solving difficulties experienced by the ASD children in the present study could, therefore, reflect ToM deficits. Weakened central coherence, a cognitive style frequently observed in ASD (Frith, 2003), may also have contributed to some of the problem solving difficulties evident in the ASD children. Central coherence enables a person to integrate information to derive the full meaning or gestalt (see e.g., Jolliffe & Baron-Cohen, 2000; Noens & van Berckelaer-Onnes, 2005). Numerous aspects of a problem are presented (pictorially or verbally) in both problem solving tests used in the current study, and difficulties with integrating information and gaining a coherent understanding of the various aspects of the story presented may have reduced the ability to effectively make sense of the situation presented in the scenarios or stories. Thus, although the findings of the current study extend the clinical

127 understanding of the extent of difficulty experienced by children with ASD, AS, or HFA with problem solving in social contexts, applying this knowledge to the development of intervention programmes is problematic, given the multiple skills that may contribute to successful problem solving. 5.8

Specific Limitations and Future Research Directions

The present study has a number of specific limitations that reduce the clinical utility of the findings. The investigations undertaken in the preceding chapters of this thesis required that the ASD participants not be matched to control participants on verbal ability. The tests of problem solving used in the present study are language-based. Core language skill was used as a covariate in the non-differentiated ASD analysis only. Apart from language skill, variables such as life experience, ToM, central coherence, and executive function skill were not controlled in the study. The influence of these variables, therefore, could not be tested. Additionally, although the scoring criteria from the TOPS-E, R (Zachman et al., 1994) and the TOPS-A (Bowers et al., 1991) instruction manuals were strictly applied, inter-rater reliability on scoring decisions is needed. Future research directions include replicating the present study with a larger sample. The finding that differentiation between AS and HFA approached significance requires further investigation. Additionally, studies investigating the interplay between problem solving and cognitive attributes such as ToM and central coherence, and executive function skills are warranted. 5.9

Conclusion

Clinically relevant research findings provide an evidence-base for treatment and support (Rutter, 2005). The theoretical approach adopted in the present study, which viewed the skills required for social problem solving as diverse yet interrelated, precludes the translation of research findings into therapy goals. Nonetheless, the study’s findings increase the clinical understanding of the extent of difficulty associated with social problem solving in ASD by supporting and extending Griswold et al.’s (2002) findings. In the present study, the non-delineated ASD group of children had compromised social problem solving skills relative to the control group

128 when the effect of language was removed. The social problem solving skills within the non-delineated ASD group of children were revealed as being more homogeneous than their typically developing peers. The scoring systems of the assessments used in the study, however, mask the actual spread of scores. There was no significant difference between the reclassified AS and HFA groups of children on social problem solving, which supports the trend in the findings from the validation studies in the preceding chapters of this thesis. Section 1.4.1 of this thesis argued one limitation of research undertaken with the ASD population is the restricted range of assessments used to investigate language and communication skills. The studies reported thus far have aimed to address this methodological limitation by examining semantic development, morphological and syntactic content, meaning, language structure, and social communication skills such as pragmatics, abstract language, and socially based problem solving skills in children presenting with AS, HFA, or ASD. An additional limitation identified in the literature review is the present research focus on children with the disorder (refer to section 1.4.2). Chapters 6 - 8 aim to address this limitation by reporting on a series of studies focusing on the language and communication skills of a group of adults with AS, HFA, or ASD.

129

6

Adults with Autism Spectrum Disorder: Linguistic and Pragmatic Skills 4

6.1

Introduction

ASD is a life-long developmental disability with a neurological basis (Hill & Frith, 2003). Hence, it is a disorder of development, as opposed to a disorder of childhood, and some individuals may reach adulthood before a diagnosis is given (Frith, 2003). As ASD has often been viewed as a childhood disorder, it is, therefore, not surprising that there is a lack of awareness of the implications of the disorder for adults (Tantam, 2003). While interest in ASD has surged in recent years, relatively little is known about the needs of adults with the disorder (Howlin, 1997; Tantam, 2003). Language and communication difficulties are central to the autistic presentation (APA, 1994), and language and communication difficulties contribute directly and indirectly to many of the behavioural, social, and academic outcomes for children with an autistic diagnosis (Lord, 2000). In addition, language and communication difficulties may be impediments to successful psychosocial and vocational outcomes in adults with ASD (Howlin, 1997). Studies of adults with a history of childhood developmental language disorders (Clegg, Hollis, Mawhood, & Rutter, 2005; Eales, 1993) have suggested that childhood language difficulties may negatively impact upon psychosocial wellbeing in adulthood. Clegg et al., for instance, investigated psychosocial outcomes for adults with a history of childhood developmental language disorders. They found that poor vocational outcomes, such as unstable work histories or long periods of unemployment, and deficits in social

4

“Linguistic and pragmatic language skills in adults with Autism Spectrum Disorder: A pilot study,”

by F. M. Lewis, G. C. Woodyatt, and B. E. Murdoch, in press, Research in Autism Spectrum Disorders, was adapted from this chapter.

130 relationships and close friendships were associated with the adults with a history of language impairment. Moreover, Eales investigated adults with a history of developmental receptive language in his study. He likewise described negative psychosocial consequences of early language impairment, in that pragmatic language deficits contributed to social impairments in the adults with a developmental history of receptive language impairment. If language difficulties reduce psychosocial outcomes for adults with histories of language disorder, but not necessarily ASD, an accurate clinical picture of the language difficulties of individuals with ASD is, therefore, needed to determine if specific language intervention and support is warranted for this population. Studies investigating basic language skills in adults with ASD have identified only mild linguistic difficulties (Rumsey & Hamburger, 1988), and some restrictions in receptive and expressive semantic knowledge (Howlin, 2003). Findings by Rumsey and Hamburger and Howlin, however, may be over-estimating the basic language skills in adults with ASD. Their studies investigated a narrow range of language abilities only, thus possibly contributing an inaccurate clinical picture of the day to day language skills of adults with disorders on the autism spectrum. Howlin (1997), for instance, has proposed that receptive language deficits in adults with the disorder may reduce vocational success. Receptive language deficits were identified in the children with ASD involved in an earlier study in this thesis (see chapter 2). Others (e.g., Koning & Magill-Evans, 2001) have likewise described receptive language difficulties in children with an autism disorder. The findings reported in chapter 2 of this thesis, and those reported by Koning and Magill-Evans were based on performance on the CELF-4 (Semel et al., 2003) and the Clinical Evaluation of Language Fundamentals- Revised (CELF-R) (Semel, Wiig, & Secord, 1987) respectively. Both test instruments assess a range of receptive language skills such as following increasingly complex oral directions, understanding spoken paragraphs, word classes, and semantic relationships, yet Howlin’s (2003) assessment of receptive skills in adults with ASD was restricted to the British Picture Vocabulary Scale (Dunn et al., 1997; Dunn et al., 1982), an assessment of receptive semantic skills only. Overall receptive language skills may not necessarily be determined by

131 receptive semantic skills alone. The limited data regarding specific language difficulties in adults with ASD may likewise be contributing to an inaccurate clinical picture. Rumsey and Hamburger’s (1988) finding of mild linguistic difficulties was based on performance on subtests of the Neurosensory Center Comprehensive Examination for Aphasia (Spreen & Benton, 1969) and the Boston Naming Test (Kaplan, Goodglass, & Weintraub, 1978). No combination of tests within the test battery, however, was identified as more discriminating than others. No specific deficits could, therefore, be identified. An additional limitation to the current clinical knowledge regarding the language skills associated with the ASD presentation in adulthood is that previous studies have documented that the skills within ASD are not homogenous (Emerich et al., 2003; Mawhood et al., 2000; Rumsey, 1985; Rumsey & Hamburger, 1988), yet no studies to date have undertaken within-group analysis to investigate the heterogeneity of language skills of adults on the autism spectrum. 6.2


Childhood language difficulties may reduce psychosocial wellness in adulthood (Clegg et al., 2005; Eales, 1993). Impaired language and communication skills are features central to the ASD presentation (APA, 1994), and as presented in chapters 2 and 4 of this thesis, numerous studies have described linguistic (e.g., Koning & Magill-Evans, 2001; Mayes & Calhoun, 2001; Prior et al., 1998; Shields et al., 1996; Szatmari et al., 1995) and pragmatic deficits (e.g., Dennis et al., 2001; Emerich et al., 2003; Losh & Capps, 2003; Paul et al., 2005; Shields et al.; Shriberg et al., 2001) in children and adolescents with ASD. Seltzer et al. (2003) have suggested that overall use of language and nonverbal communication skills in ASD improves from childhood to adolescence, and that the pattern of improvements continues into adulthood. Howlin (1997), however, contends that most adolescents and adults with the disorder continue to exhibit language difficulties, and believes that improved services for adults are needed if long-term outcomes are to be enhanced (2003). Seltzer et al.

132 (2003) concur, suggesting that adults with ASD have no lesser need for services and support than children with the disorder. It is possible that support services are needed in the area of language and communication in the adult ASD population, as skill deficits in these areas have been proposed as impediments to successful psychosocial and vocational outcomes in adults with ASD (Howlin, 1997). Determination of the type and level of support offered to adults with ASD requires a clinical appreciation of their communicative needs, but generalising the results from child and adolescent studies to adults should be avoided as language skills in ASD continue to change from adolescence to adulthood (Seltzer et al.) and the maturational changes in language skill from childhood to adulthood are not, as yet, clearly understood in the disorder (Howlin, 1997). Few studies to date have focused on the language and communication skills in adults with ASD ( Howlin, 2003; Rumsey & Hamburger, 1988), and the current clinical appreciation of the everyday, and hence, vocational communication needs of adults with the disorder is limited due to restricted methodologies applied in the previous research studies. An assessment of a comprehensive range of linguistic and pragmatic language functions may provide clinically useful findings on the language and communication status of adults with ASD. 6.3


Seltzer et al. (2003) have argued that developmentally appropriate services are needed for adults with ASD diagnoses. To offer appropriate support, an accurate clinical understanding of their language and communication strengths and weaknesses is needed. Hence, the current study investigated the linguistic, pragmatic language, and pragmatic communication skills in a group of adults with the diagnosis relative to typically developing peers. Additionally, to provide clinically relevant information, hierarchical cluster analysis of the ASD participants was performed to profile the heterogeneity of language skills. As previous researchers (Howlin, 2003; Rumsey & Hamburger, 1988) have described minor linguistic impairments in adults with ASD, it was hypothesised that there would be significant differences between the two groups of adults on a range of linguistic and pragmatic language measures. As communication deficits, including deficits in social communication, are central to the

133 autistic presentation (APA, 2000), it was also hypothesised that there would be a significant difference between the two groups on pragmatic communication skills. It was further hypothesised that subgroups within the ASD group, based on linguistic and pragmatic language skill, would be evident. 6.4

Method

6.4.1 Participants Seventeen adults with average intelligence [as evidenced by no previously documented intellectual impairment] (8 male, 9 female; M age: 34 years 8 months; SD: 12 years 2 months; range: 18 years – 67 years; M years of schooling: 12; range: 8 years – 15 years) and a diagnosis of AS, HFA, and/or ASD were recruited through the Asperger Syndrome Support Network of Queensland, Autism Queensland, and newspaper articles. Table 6.1 displays the demographic details of the participants. In addition to a diagnosis of AS/HFA/ASD, inclusion criteria for the study were English as a first language, no neurological disease or trauma, no history of drug or alcohol abuse, no other co-morbid condition, and normal vision and hearing. Due to financial constraints, no formal IQ testing was undertaken for the study. Documentation of diagnosis was requested before the commencement of testing. Diagnoses made by psychiatrists and psychologists were accepted for inclusion into the study. As mentioned in section 2.4.1, there have been numerous criteria for diagnosing AS (e.g., APA, 1994; Gillberg & Gillberg, 1989; Szatmari et al., 1989; WHO, 1992). The criteria used by the various professionals for diagnosis of the adult participants investigated in this thesis were not known.

134

Table 6.1 ASD adult participant demographics ID

Sex

Age in years

Ed level a

(M: 8; Fe: 9)

When

Dev lge hist

Diag (Source)

diagnosed

Current

Early

employment

intervention

status 1

Fe

18

Sec

Child

Delayed

HFA (psychiat)

Supported training S P programme

2

M

29

Sec

Adult

Delayed

AS (psychiat)

Unemployed

SP

3

M

27

Tert

Adult

Unsure

AS (psych)

Full-time

Nil

4

M

28

Sec

Child

Delayed

HFA (psychiat)

Unemployed

S P and Beh

5

M

48

Tert

Adult

Unsure

AS (psychiat)

Full-time

Nil

6

M

29

Post-Sec

Adult

Normal

AS (multi-

Unemployed

SP

disciplinary team: Autistic Centre 7 8

Fe Fe

30 42

Post-Sec Sec

Adult Child

b

Unsure

AS (psychiat)

Unemployed

Nil

Delayed

Autism (psych)

Part-time by

Beh

choice Table continues


Sex

Age in years

Ed level a

(M: 8; Fe: 9)

When

Dev lge hist

Diag (Source)

diagnosed

Current

Early

employment

intervention

status

9

Fe

31

Tert

Adult

Unsure

AS (psych)

Full-time

Nil

parenting 10

M

30

Tert

Adult

11

Fe

22

Post-Sec

Adult

12

Fe

26

Sec

Adult

13

M

67

Post-Sec

Adult

c

Unsure

AS (psych)

Full-time

Nil

Delayed

AS (psych)

Unemployed

SP

Unsure

AS (psych)

Unemployed

Beh

Unsure

AS (psych)

Retired from full-

Nil

time work 14

Fe

38

Tert

Adult

Unsure

AS (psychiat)

Part-time by

Nil

choice 15

Fe

32

Sec

Adult

Delayed

AS (psychiat)

Unemployed

Nil

16

M

48

Post-Sec

Adult

Unsure

AS (psych)

Unemployed

Nil

17

Fe

48

Prim

Adult

Delayed

AS (psych)

Unemployed

SP Table continues

135

136

Table 6.1 continued Note. ASD = Autism spectrum disorder. AS = Asperger syndrome. HFA = High functioning autism. Ed level = Educational level. Dev lge hist = Developmental language history. Normal = Language onset prior to age two years. Delayed = Language onset after two years of age. Unsure = Developmental language history not available. Diag = Independently determined diagnosis. Source = Source of diagnosis. Early intervention = Type of intervention provided during childhood. S P = Speech pathology intervention. Beh = Behavioural intervention. Prim = Primary. Sec = Secondary. Post-Sec = Post-secondary vocational trade. Tert = Tertiary. Psych = Psychologist. Psychiat = Psychiatrist. M = Male. Fe = Female. a

= Primary education consisted of 8 years of education; Secondary consisted of 12 years of education; Post-secondary vocational

consisted of 11 years of education including trade qualifications; Tertiary consisted of 15 years of education. b = Diagnosed as autistic at three years of age by a psychologist, then diagnosed as an adult as AS by a psychiatrist. c = Diagnosed in childhood with a number of psychiatric disorders, then diagnosed in adulthood as AS.

137 Thirteen adults (6 males, 7 females; M age: 34 years 7 months; SD: 13 years 6 months; range: 18 years – 65 years; M years of schooling: 13; range: 11 years – 15 years) with non-significant developmental histories were recruited as control participants. Student t tests indicated there were no significant differences between the control group and the ASD group on age (t = 0.040; p =.968), nonverbal cognitive skill (t = -1.765; p =.857), or years of schooling (t = -1.322; p =.197). Recruitment of control participants was through newspaper articles in local papers, and local sporting clubs in the south-east corner of the state of Queensland, Australia. To be included in the control group, participants were required to have English as a first language, no history of neurological trauma or disease, no history of drug and/or alcohol abuse, and normal hearing and vision. In addition, to avoid co-morbidity and family links with autism and language difficulties, there was to be no family history of autism, and no developmental or family history of language delays or difficulties. Of the 13 control participants, 10 were engaged in full-time employment, one was a full-time parent, one had retired after life-long full-time work, and one was a full-time university student. 6.4.2 Measures The present study is part of a larger study examining the clinical presentation of adults with ASD and adults with right- and left- hemisphere focal brain injury. The language assessments administered, therefore, needed to be accessible to the brain-injured population. All participants were assessed using the Western Aphasia Battery (WAB) (Kertesz, 1982), the RHLB (Bryan, 1989), Form A of the TONI-2 (Brown et al., 1990), and the Rating Scale of Pragmatic Communication Skills (RSPCS) (Burns, Halper, & Mogil, 1985). The WAB and the RHLB were chosen to assess linguistic and pragmatic language skills respectively. The RSPCS was chosen as a subjective measure of pragmatic social skills. The rationale for the inclusion of the TONI-2 is presented in section 2.4.2. For the purposes of the current study, only the language component of the WAB (Kertesz, 1982) was administered. The WAB subtests administered were: Spontaneous Speech (Information and Fluency), Auditory Verbal Comprehension, Repetition, and Naming. Spontaneous Speech measures functional communication,

138 and is designed to elicit conversational speech from the participant through an interview and picture description. The Information score is based on the accuracy, the completeness of content, and the relevance and appropriateness of responses. The Fluency score reflects grammatical and syntactic aspects of language production. Auditory Verbal Comprehension consists of an assessment of increasingly complex comprehension tasks. The tasks include Yes/No questions, single word recognition, and the assessment of comprehension of sequential commands, where the length of the sentence and the number of clauses within the sentence increases. Repetition assesses the ability to repeat high frequency single words, composite words, high and low probability sentences and sentences of increasing length and grammatical complexity. Naming assesses competence on confrontation naming of objects, category naming, sentence completion, and responsive speech. The WAB (Kertesz, 1982) subtests can be scored individually, as well as combined to provide a WAB Aphasia Quotient (WAB AQ), an index of overall linguistic skill. The WAB is standardised for the adult population, but is intended to be used for adults with brain injury. It was included in the present study as no broadbased language assessment, such as CELF-4 (Semel et al., 2003), has been designed for the non-neurologically impaired adult population. The psychometric properties of the WAB (Kertesz, 1982) are adequate (Shewan & Kertesz, 1980). The correlation coefficients for the internal consistency (.97), test-retest reliability (at or above .90 for all WAB subtests used to calculate the WAB AQ), and inter-rater reliability (.90 for all subtests except Fluency) suggest the WAB is a reliable measure. When matched for content to the Neurosensory Center Comprehensive Examination for Aphasia (Spreen & Benton, 1969), the correlation coefficient (.97) suggests the WAB has a high degree of construct validity. Aspects of the RHLB (Bryan, 1989) relevant to the present study are presented in section 4.4.2. The RHLB manual includes a table to convert adult raw scores to relative T Scores for each of the subtests. Examples of the subtest items are found in Appendix E. The RSPCS (Burns et al., 1985) is a subjective assessment that covers 12

139 pragmatic communication skills. The 12 pragmatic skills are presented in four pragmatic sub-sections. Nonverbal Communication includes the use of intonation, facial expression, eye contact, and gestures and body proxemics, Conversational Skills encompasses conversation initiation, turn-taking, and verbosity, Use of Linguistic Context addresses topic maintenance, presupposition, and referencing skills, while Organisation of a Narrative examines organisation and completeness of discourse. The RSPCS is shown in Appendix I. A Total RSPCS score is obtained by combining the ratings from each of the 12 pragmatic skills, giving a maximum possible score of 60. Psychometric properties of the RSPCS are not provided with the rating scale. 6.4.3 Procedure All ASD participants were requested to complete a questionnaire prior to commencement of testing. The questionnaire was completed by participants, parents, or carers. As well as providing identifying data, the questionnaire requested a retrospective history of language development. Developmental language history was recorded with the aim of investigating whether there were quantitative differences in language outcomes for those with and without delayed onset of language, as per the validation studies (refer to section 1.3.1). The use of the questionnaire for the reclassification of ASD participants was discussed in section 2.4.3, and the questions pertaining to developmental language history are displayed in Appendix C. Inter-rater reliability of 100 % was achieved for determining whether language onset was normal or delayed on 54 % of the total child and adult data set for the thesis (i.e., 20/37 developmental language histories). (Refer to section 2.4.3 for details). Forty-seven percent (8/17) of the adult participants were unable to provide details of their developmental language history. The unreliability of reported developmental language history has been discussed in section 2.6.1. The available data set on developmental language history was judged to be too unreliable to undertake an examination of categorical distinctions based on language history for the adults with ASD involved in the study. Therefore, for this and subsequent studies involving the adult participants, the theoretical approaches of combined and spectrum studies only were applied. Scores from the WAB-AQ (Kertesz, 1982) and the six

140 RHLB (Bryan, 1989) subtests were used for the hierarchical cluster analysis undertaken for the spectrum study. For the purposes of the current study, each participant completed a RSPCS (Burns et al., 1985) (RSPCS-SR). A Significant Other also completed the rating scale on the participant (RSPCS-SO). Significant Others were spouses, siblings, parents, friends, or carers. Participants and their Significant Others were given no opportunity to collaborate on completing the RSPCS. The Total RSPCS-SR and the Total RSPCSSO scores were used for the analyses. Two RSPCS data were not obtained. One female with ASD stated that she was not capable of rating her own pragmatic skills and one female control stated that she was unwilling to allow a Significant Other to profile her skills. The Behavioural and Social Sciences Ethical Review Committee at the University of Queensland, Australia gave ethical approval to the study involving the adult participants. Signed informed consent was required from the participants prior to acceptance into the study. All testing was undertaken at the University of Queensland or in the home of the participant. Each assessment was administered in a distractionfree environment in a standardised manner according to the instruction manual. No time limits were placed on participant responses unless particular subtests required time limits. 6.5

Results

6.5.1

Performance by the Adults on the WAB, the RSPCS-SR, the RSPCS-SO, and the RHLB

Means and standard deviations for all measures are listed in Table 6.2. For significance at the 0.05 level, p must be equal to or less than 0.004, as determined by the Bonferroni adjustment to correct for the multiplicity of tests (Price, 2000). No significant difference between the two groups was found on mean TONI-2 performance (p ≥ .004). For each measure, a Levene’s test was carried out, and where necessary, appropriate corrections for non-homogeneity of variance were undertaken.

141 The student t test revealed significant differences between the two groups, with the ASD group being less proficient on the WAB AQ (p ≤ .001), Auditory Verbal Comprehension (p ≤ .004), Naming (p ≤ .004), Comprehension of Inferred Meaning (p ≤ .001), Appreciation of Humour (p ≤ .004), Production of Emphatic Stress (p ≤ .001), the RSPCS-SR (p ≤ .001), and the RSPCS-SO (p ≤ .001). All other measures failed to reach significance (p ≥ .004 on all measures).

142 Table 6.2

Descriptive statistics of the performance on the WAB, the RHLB, the RSPCS-SR, and RSPCS-SO by the adult participants ASD

Control

(N = 17)

(N = 13)

t

Sig. (2tailed)

Variable

M

SD

M

SD

TONI-2

91

14

99

11

-1.765

.857

WAB AQ

95.90

3.40

99.50

1.10

-4.130

.001 ***

Inform

9.94

.24

10.00

.00

-0.871

.391

Fluency

9.41

1.12

10.00

.00

-2.163

.046

Comp

9.49

.57

9.99

.04

-3.571

.003 **

Rep

9.58

.42

9.79

.53

-1.210

.237

Naming

9.52

.55

9.98

.08

-3.366

.004 **

MP

52.80

9.99

60.69

.75

-3.258

.005

WM

53.71

7.33

57.38

3.07

-1.866

.075

CIM

41.24

7.24

52.15

4.06

-5.234

.001 ***

AH

46.88

12.79

58.08

6.30

-3.144

.004 **

LS

57.29

8.08

62.00

.00

-2.403

.029

45.71

6.23

55.85

2.19

-6.223

.001 ***

9

52

7

-4.679

.001 ***

6

-5.334

.001 ***

RHLB

PES a

RSPCS-SR (60) a

RSPCS-SO (60)

b

37 (16) 38

8

b

53 (12)

Note. WAB = Western Aphasia Battery. RHLB = Right Hemisphere Language Battery. TONI2 = Test of Nonverbal Intelligence-2nd Edition. ASD = Autism spectrum disorder. WAB AQ = Aphasia Quotient from the Western Aphasia Battery. Inform = Information. Comp = Auditory Verbal Comprehension. Rep = Repetition. M P = Metaphor Picture Test. W M = Written Metaphor Test. C I M = Comprehension of Inferred Meaning. A H = Appreciation of Humour. L S = Lexical Semantic Test. P E S = Production of Emphatic Stress. RSPCS-SR = Rating Scale of Pragmatic Communication Skills-Self-Rating. RSPCS-SO = Rating Scale of Pragmatic Communication Skills-Rating by a Significant Other. a

= Maximum possible score. b = Missing data. The number preceding b indicates the number of

cases involved in the analysis. ** = p ≤ .004 (Bonferroni adjustment). *** = p ≤ .001.

143 Descriptively, the standard deviations for the ASD group are greater than the standard deviations for the control group on all measures except Repetition, suggesting greater score dispersal for the ASD group on all other performance measures (refer to Table 6.2). Figures 6.1 - 6.9 display the score dispersal of the ASD and control adults’ performance scores on the WAB AQ, the six RHLB subtests, the RSPCS-SR, and the RSPCS-SO. Outlier scores were noted for the ASD participants on Lexical Semantic Test and Production of Emphatic Stress. Outliers from the control participants were noted on the WAB AQ and the RSPCS-SR, and far outliers on the WAB AQ, Metaphor Picture Test, Written Metaphor Test, and Production of Emphatic Stress. There is greater spread in the scores for the ASD group compared to the control group on the WAB AQ, the Metaphor Picture Test, Written Metaphor Test, Comprehension of Inferred Meaning, Appreciation of Humour, Lexical Semantic Test, Production of Emphatic Stress, the RSPCS-SR, and the RSPCS-SO.

144

102

100

WAB- 98 AQ

12

96

7

94 92

90 88

N=

17

13

ASD

Control GROUP

Figure 6.1.

Box plots displaying the dispersal of scores from the adult ASD

and control groups’ performance on the Western Aphasia Battery Aphasia Quotient.

145

60

7 16

MP

50

40

30

20

N=

17

13

ASD

Control GROUP

Figure 6.2.


and control groups’ performance on the Metaphor Picture Test.

146

70

60

WM 50

4 7 9

40

30

N=

17

13

ASD

Control GROUP

Figure 6.3.


and control groups’ performance on the Written Metaphor Test.

147

60

50

CIM

40

30

20

N=

17

13

ASD

Control GROUP

Figure 6.4.


and control groups’ performance on Comprehension of Inferred Meaning.

148

70

60

AH

50

40

30

20

N=

17

13

ASD

Control GROUP

Figure 6.5.


and control groups’ performance on Appreciation of Humour.

149

70

60

LS 50

40

15

30

6 N=

17

13

ASD

Control GROUP

Figure 6.6.


and control groups’ performance on the Lexical Semantic Test.

150

50

5 8 9

PES

40

2 30

N=

17

ASD

Figure 6.7.

13

GROUP

Control


and control groups’ performance on Production of Emphatic Stress.

151

70 60

50

RSPCS-SR 40

2 30

20 10

N =

16

13

ASD

Control GROUP

Figure 6.8.


and control groups’ performance on the Rating Scale of Pragmatic Communication Skills-Self Rating.

152

70

60

50

RSPCS-SO 40 30

20

10

N=

17

12

ASD

Control GROUP

Figure 6.9.


and control groups’ performance on the Rating Scale of Pragmatic Communication Skills-Significant Other Rating.

As shown in Table 6.2, the ASD group’s RSPCS-SR mean was similar to the ASD group’s RSPCS-SO mean. The control group’s RSPCS-SR and RSPCS-SO means were also similar. 6.5.2

Clusters of Adults with ASD Based on Performance on the WAB AQ and RHLB

To investigate potential subgroupings within the ASD group, performance scores from the WAB AQ and the six RHLB subtests were entered for an agglomerative hierarchical cluster analysis. The hierarchical clustering of the ASD adults is shown in Figure 6.10.

153

Figure 6.10.


adults based on performance on the Western Aphasia Battery Aphasia Quotient and the six subtests from the Right Hemisphere Language Battery.

Two subgroups can be identified from the analysis. The resultant subgroups are descriptive only. Table 6.3 displays the membership of the subgroups. Each participant’s score is defined as within a number of standard deviations from the control group’s mean. Performance scores within each subgroup support the partitioning decision. Subgroup 1 is characterised by mild to moderate overall linguistic difficulties, generally intact metaphorical language skills, mild to moderate difficulty with inferential language and appreciation of humour, intact lexicalsemantic flexibility, and moderate-severe difficulties using emphatic stress to convey meaning in speech. Subgroup 2 is characterised by generally severe overall linguistic difficulties, and moderate-severe difficulties with metaphorical and inferential language, appreciation of humour, lexical-semantic flexibility, and the production of emphatic stress to convey contrastive meaning in speech.

154

Table 6.3

Subgroups of adults with ASD based on performance on the WAB AQ and the six RHLB subtests

Subgroup

ID

WAB AQ

1 (n = 10)

5

+1


14

(SD: 14, 7) M TONI-2: 95.9: (SD: 10.4)

MP

WM

CIM

AH

LS

PES

0

+1

-1

-1

+1

-2

-1

-2

+1

-2

-1

+1

-3

16

-1

-2

+1

-3

-2

-6

+1

10

-2

-2

+1

-1

-1

-11

-4

1

-4

0

-2

-1

-2

+1

-7

7

-2

0

+1

-3

-3

+1

-7

11

-2

0

+1

-4

-1

+1

-8

3

-2

0

+1

0

+1

+1

-4

8

-2

0

+1

-1

+1

+1

-8

13

+1

-2

-2

-2

+1

+1

-2

Table continues


ID

WAB AQ

MP

WM

CIM

AH

LS

PES

2 (n = 7)

9

-8

-20

+1

-3

+1

+1

-3

M age: 31, 8 (SD: 7, 4)

2

-7

-16

-4

-6

-5

+1

-12

M TONI-2: 84.1 (SD: 15.7)

17

-10

-20

-4

-5

-4

+1

-10

4

-6

-16

-2

-3

-5

-6

-4

12

-8

-12

-2

-5

-4

-6

-5

15

-3

-16

-4

-3

-5

-19

-5

6

-4

-32

+1

-6

-5

-26

-5

Note. ASD = Autism spectrum disorder. WAB AQ = Western Aphasia Battery Aphasia Quotient. RHLB = Right Hemisphere Language Battery. TONI-2 = Test of Nonverbal Intelligence-2nd Edition. M P = Metaphor Picture Test. W M = Written Metaphor Test. C I M = Comprehension of Inferred Meaning. A H = Appreciation of Humour. L S = Lexical Semantic Test. P E S = Production of Emphatic Stress. Yrs, mths = Years, months. Scores are entered as within standard deviations above or below the control mean. + = Above the control mean. - = Below the control mean. Scores entered as 0 indicate no deviation from the control mean.

155

156 Non-parametric tests (Mann-Whitney U tests) established there were no significant differences between the two ASD subgroups on the TONI-2 (Z = -1.661; p = .097) and age (Z = -.588; p = .557). 6.6

Discussion

The present study investigated the linguistic and pragmatic skills of a group of adults with a diagnosis of ASD. The findings suggest that, despite presenting with comparable nonverbal cognitive skills, the adults with ASD in the present study experienced difficulties in a number of linguistic and pragmatic language areas relative to their typical peers. The ASD group’s pragmatic communication skills were rated as less proficient than the control group by the ASD participants themselves and by their Significant Others. Moreover, an examination of performance differences within the ASD participants revealed the language skills associated with the disorder ranged from average performance to severe difficulties. 6.6.1

Linguistic and Pragmatic Skills in Adults with ASD

Previous researchers have described only mild linguistic difficulties such as expressive and receptive semantic deficits in adults with ASD (Howlin, 2003; Rumsey & Hamburger, 1988). Extending previous findings, the current study determined that basic linguistic skills, such as auditory verbal comprehension and naming abilities, as assessed by the WAB (Kertesz, 1982), were compromised in the adults with the diagnosis. Additionally, the ASD participants were less competent than their peers on a range of pragmatic language skills, such as understanding inference, appreciating humour, and producing emphatic stress. Two RSPCS (Burns et al., 1985) measures were used in the analysis to provide subjective measures of social pragmatic skills, which incorporate the use of intonation, facial expression, eye contact, conversational skills, such as initiation, turn-taking, verbosity, topic maintenance, presupposition, and referencing. Ratings of pragmatic competence were significantly lower for the ASD group than the control group for both

157 the self-rating and the rating by a Significant Other. There was concordance between the self-ratings and the Significant Other ratings for both participant groups. The similar ratings of pragmatic competence by the ASD group of participants and their Significant Others suggest the ASD group possessed a relatively accurate self-perception of pragmatic skills. Previous studies of adults with developmental histories of language impairment have identified language and communication difficulties as impediments to positive psychosocial and vocational outcomes (Clegg et al., 2005; Eales, 1993). It is possible, therefore, that the linguistic and pragmatic deficits identified in the current study may be impediments to positive psychosocial and vocational outcomes in the lives of the adults with ASD. Fifteen adults with ASD (excluding one participant who was retired and one engaged in full-time parenting) and 10 control participants (excluding one full-time parent, one retired after life-long full-time work, and one in full-time university study) were in a position to engage in gainful employment. All 10 control participants were employed. Five ASD participants were employed. Two elected to work part-time. The remaining three ASD participants employed were tertiary qualified and working in their field of study. Most of the remaining 10 ASD adults expressed a desire to secure meaningful employment. Further research is needed to determine if linguistic and pragmatic deficits are impediments to obtaining gainful employment in the adults with ASD. The need for thorough and specific language assessment of adults with a diagnosis of ASD has been highlighted by the present study. A comprehensive language assessment may provide clinically relevant data that can be utilised to determine the focus of support and intervention. However, heterogeneity of linguistic and pragmatic language skills within the ASD group was evident, but not further examined, in the between-group analysis. The clinical utility of findings from this level of analysis may, therefore, be limited.

158 6.6.2

Heterogeneity of Linguistic and Pragmatic Language Skills of Adults with ASD

The description of language subgroups within the ASD group of adults supports previous findings of non-homogeneity of skills within the ASD population (Emerich et al., 2003; Mawhood et al., 2000; Rumsey, 1985; Rumsey & Hamburger, 1988). Variations in the normal population may account for some, but not all, of the heterogeneity observed in the current study. One subgroup possessed reasonably intact metaphorical language skills, experienced mild to moderate difficulties on overall linguistic tasks, and moderate-severe difficulties on the use of emphatic stress to convey contrastive meaning in speech. All three ASD participants engaged in full- time employment, and the retiree from full-time employment, were placed within this subgroup, suggesting their linguistic/pragmatic skills were not barriers to employment. Within this subgroup, however, were three adults with ASD who have faced ongoing barriers to gaining employment, suggesting factors other than linguistic and pragmatic skills, may be contributing to their poor vocational outcomes. The second subgroup experienced moderate to severe difficulties across all tasks. No ASD participant in this subgroup was engaged in employment (full-time or otherwise). Six of the ASD participants investigated in the study received speechlanguage intervention in childhood. Despite this, performance by four of the six ASD adults with a history of language intervention placed them in the subgroup characterised by moderate-severe difficulties on all tasks, suggesting they have considerable ongoing language needs. An examination of within-group differences in linguistic and pragmatic skills of the adults with ASD may provide important information that can be applied in speech pathology service delivery, as the findings of the present study strongly suggest that individuals with ASD may have differing communicative needs. Mawhood et al. (2000) have argued that examining the cause of the heterogeneity within ASD may be a means of gaining insight into the disorder. The influence of cognitive skill on the heterogeneity of language skills has previously been examined (e.g., Szatmari et al., 1995). In the present study, nonverbal cognitive skill was not a factor in the subgroup formation for the adults with ASD. This finding is inconsistent with the

159 findings presented in chapters 2 and 4 from investigations of nonverbal cognitive influences in the formation of language subgroups based on the linguistic and pragmatic language skills of the children with ASD. For the children, nonverbal cognitive skill was a factor in the formation of subgroups based on linguistic skills (see section 2.6.3), but not in the formation of subgroups based on pragmatic language skills (see section 4.6.3). The inconsistent findings between the adult and child studies in this thesis could be attributed to the different assessments used to determine the subgroups. While the same assessment (RHLB) (Bryan, 1989) was administered to assess pragmatic language skills in both the child and adult studies, the linguistic assessment used for the child study was the CELF-4 (Semel et al., 2003), which is a more comprehensive linguistic assessment than the WAB (Kertesz, 1982). Moreover, scores from the CELF-4 and the RHLB were independently entered for the cluster analyses for the child studies, whereas in the present study, the WAB AQ and the RHLB scores were entered together for the one cluster analysis. Both factors may have influenced the findings relating to the nonsignificance of nonverbal cognitive skill in the formation of subgroups in the present study. 6.7

Specific Limitations of the Study The current findings were based on performance by adults with ASD on tests

designed for the neurologically impaired population. An investigation using tests designed for the non-neurologically impaired population may provide different results. There are, however, few comprehensive linguistic and pragmatic assessments designed for adults without neurological disorders. The CELF-4 (Semel et al., 2003), a comprehensive language assessment designed for the non-neurologically impaired paediatric population, was chosen for the language assessment in the child study (refer to chapter 2). Although the CELF-4 provides Standard Scores for ages ranging from 5 years 0 months to 21 years 11 months, a number of the test items, such as the picture prompts in the Formulated Sentences subtest, may not be age-appropriate for testing an adult population. Nonetheless, an investigation using tests designed for non-neurologically impaired adolescents, such as the Test of Adolescent Language-Third Edition (Hammill,

160 Brown, Larsen, & Wiederholt, 1994) may be more age-appropriate and may provide more detailed analysis of the linguistic capabilities in adults with ASD. As mentioned in section 4.7, the lack of data regarding the psychometric properties of the RHLB (Bryan, 1989) is a major limitation of the present study. The RSPCS (Burns et al., 1985) provided useful subjective measures of social pragmatic skills. Objective measures of these skills would complement the present study. The Discourse Analysis from the RHLB (Bryan, 1989) (not undertaken in the present study) examines two-way interaction, and may provide more information regarding social pragmatic skills in the adults with ASD. (Refer to section 4.7 for a summary of the RHLB Discourse Analysis). Studies incorporating behavioural observation of video data are needed to extend the findings from the RSPCS. Further research is needed to determine if the level of self-perception of pragmatic competence noted in the ASD group works favourably in clinical outcomes for pragmatic intervention. A high proportion of the adults with ASD investigated in the study were not employed. The relationship between their language and pragmatic deficits and employability was not examined in the present research. Further investigation is warranted to determine if the communication deficits identified in this cohort of adults with ASD are barriers to meaningful employment. 6.8

Conclusion

Previous findings have overestimated the language skills in adults with a diagnosis of ASD. The adults with ASD in the current study presented with compromised skills in overall linguistic ability, auditory verbal comprehension, naming skills, understanding non-contextual language, and the use of emphatic stress to convey contrastive meaning in speech. Subjectively, the ASD group was rated significantly below the control group on social pragmatic skills such as nonverbal skills, conversational skills, and the use of linguistic context. Furthermore, the group of adults with ASD did not present with homogeneous language skills. The identification of a range of linguistic and pragmatic language strengths and weaknesses in the adults with ASD investigated in the current

161 study has clinical relevance. The majority of the ASD participants investigated in the present study, including all of the ASD participants in the least competent subgroup, were unemployed. A number of ASD-specific employment and training services are available (e.g., National Autistic Society [NAS], 2006), yet to offer support specific to the individual, an accurate understanding of their communicative needs is necessary. The findings of the present study suggest that adults diagnosed with ASD should undergo a comprehensive language and pragmatic assessment to allow for specific assistance to be tailored to their individual needs. Workplace intervention may need to include educating employers and fellow employees about possible language and pragmatic difficulties that may be impediments to successful employment for those with the diagnosis. As proposed by Manjiviona (2003) [and presented in section 1.4.1], other investigations are warranted before an accurate clinical understanding of the language and communication skills associated with ASD is gained. In order to adequately define the language and communicative needs of the adults with the disorder described in the present study, chapters 7 and 8 examine their metalinguistic and social problem solving abilities respectively.

162

7

Metalinguistic Abilities in Adults with Autism Spectrum Disorder 5

7.1

Introduction

Results presented in chapter 6 suggest previous research (Howlin, 2003; Rumsey & Hamburger, 1988) may have over-estimated the linguistic and pragmatic language skills of adults with a diagnosis of ASD. The findings of linguistic and pragmatic difficulties identified in the adults with the disorder investigated in this thesis may not, however, adequately convey the extent of language difficulty experienced by this population. Methodological limitations of studies to date, including the study reported in chapter 6, may be reducing the clinical recognition of the language difficulties associated with ASD in adulthood. Assessments such as subtests from the Neurosensory Center Comprehensive Examination for Aphasia (Spreen & Benton, 1969), the Boston Naming Test (Kaplan et al., 1978), the Expressive One Word Picture Vocabulary Test (Gardner, 1982), the British Picture Vocabulary Scale (Dunn et al., 1997; Dunn et al., 1982), the WAB (Kertesz, 1982), and the RHLB (Bryan, 1989), for instance, are designed to assess basic language skills only. Such assessments may be insensitive to some language deficits. It has been proposed, for instance, that individuals with neurological conditions may present with generally intact basic language skills as assessed by neuropsychological test batteries, but experience difficulties with language tasks that require complex cognitive-language processing (Murdoch & Lethlean, 2000). Klonoff, Clark, Oger, Paty, and Li (1991), for

5

“Communicative competence and metalinguistic ability: Performance by children and adults with Autism

Spectrum Disorder,” by F. M. Lewis, B. E. Murdoch, and G. C. Woodyatt, 2007, Journal of Autism and Developmental Disorders, 37, 1525-1538, was adapted from this chapter.

163 example, described intact functional language skills in multiple sclerosis (MS) based on performance on verbal subtests of the Wechsler Adult Intelligence Scale-Revised (Wechsler, 1981). In contrast, Lethlean and Murdoch (1997) described impaired performance in MS on understanding ambiguous sentences, metaphorical and inferential language, and the production of complex sentence constructions. Lethlean and Murdoch’s findings were based on performance on the Test of Language Competence (Wiig & Secord, 1985), a test designed to assess metalinguistic skills, or the capacity to use and interpret language in a flexible, rational, and goal-oriented manner in an ever-changing communicative environment (Wiig, 1989). 7.2


It has been proposed that language difficulties may reduce psychosocial and vocational outcomes in adults with ASD (Howlin, 1997; Howlin, Alcock, & Burkin, 2005), and studies of adults with a history of childhood developmental language disorders (Clegg et al., 2005; Eales, 1993) have found that language difficulties in childhood may negatively impact upon psychosocial outcomes in adulthood (refer to section 6.1 for details of the studies by Clegg at al. and Eales). A range of language and communication difficulties, such as auditory verbal comprehension deficits, reduced linguistic ability, difficulties with the comprehension of non-contextual language, and impaired production of contrastive stress, were identified in the previous chapter of this thesis. The findings of metalinguistic deficits in MS, a neurological disorder resulting primarily from dysfunction of the white matter of the central nervous system (Lublin & Reingold, 2003) may have implications for ASD, as it has been hypothesised that dysfunctional white matter may be central to the autistic presentation (Ellis & Gunter, 1999). An investigation of performance by adults with ASD on language tasks that require complex cognitive-language processing is, therefore, needed to provide a more clinically accurate picture of the language difficulties associated with the disorder. Additionally, findings from chapters 2 - 6 of this thesis support previous reports (Emerich et al., 2003; Rumsey, 1985; Rumsey & Hamburger, 1988) of the heterogeneity of the language skills associated with ASD. Consequently, in order to accurately reflect the

164 language skills in adults with ASD, within-group differences on performance on complex language tasks should be examined. 7.3


The current study aimed to broaden the scope of previous research into the language skills of adults on the autism spectrum. As discussed in section 6.4.3, the theoretical framework of the validation approach could not be applied in the present study due to the incomplete data set relating to the developmental language histories of the adult participants. Therefore, the first aim was to determine if performance on tasks requiring complex metalinguistic skills could differentiate a group of adults with ASD from a group of typical peers. The second aim was to examine the results of the language testing to determine if subgroups within the ASD group could be identified through hierarchical cluster analysis based on current performance on complex language tasks. It was hypothesised that the adults with ASD would present with significantly different language skills relative to their peers. As subgroups of children on the autism spectrum have previously been described on a range of language or communication skills (Prior et al., 1998; Verte, Geurts, Roeyers, Rosseel, et al., 2006), it was hypothesised that subgroups within the adult ASD group, based on metalinguistic performance, would be evident. 7.4

Methods

7.4.1

Participants

Recruitment details of the 17 adults with a diagnosis of ASD and the 13 control participants are described in section 6.4.1. The demographic details of the adult participants are displayed in Table 6.1. 7.4.2

Measures

All participants were assessed using Level 2 of the TLC-E (Wiig & Secord, 1989) and the TONI-2 (Brown et al., 1990). The TLC-E assesses metalinguistic competence, that is,

165 high-level language skills that reflect a person’s capacity to use and interpret language in a flexible manner (Wiig, 1989), and has been described in section 3.4.2. Examples of the four TLC-E subtests are presented in Appendix D. Details of the TONI-2 have been presented in an earlier section (see section 2.4.2). 7.4.3

Procedure

The TLC-E (Wiig & Secord, 1989) is not a standardised test for the adult population but was chosen for the present study as there is no equivalent standardised test for adults. Standard Scores, corrected for age, are provided for a maximum age of 18 years 11 months; therefore, the adult raw scores could not be converted to Standard Scores. Raw scores were entered for all statistical analyses. Raw scores from all four TLC-E subtests were entered for the hierarchical cluster analysis. 7.5

Results

7.5.1

Performance by the Adults on the TLC-E

Group means and standard deviations for all measures are shown in Table 7.1. For each measure, a Levene’s test was carried out, and where necessary, appropriate corrections for non-homogeneity of variance were undertaken. A level of p ≤ 0.05 was utilised for determination of statistical significance. Student t tests indicated there was no significant difference between the two groups on age (p ≥ .05), and no difference on nonverbal cognitive ability (p ≥ .05). There were significant differences between the means of the two participant groups on Ambiguous Sentences (p ≤ .05), Listening Comprehension: Making Inferences (p ≤ .05), Oral Expression: Recreating Sentences (p ≤ .01), and Figurative Language (p ≤ .01), with the ASD group less competent than the control group on all four measures.

166

Table 7.1

Descriptive statistics of the adult participants on the TLC-E ASD

Control

(N =17)

(N = 13)

Sig. (2-

t

tailed)

Variable

M

SD

M

SD

Age (yrs, mths)

34, 8

12, 2

34, 7

13, 6

TONI-2

91

14

99

Amb Sent (39 b)

25

11

List Comp (36 b)

26

Oral Exp (78 b)

0.040

.968

11

-1.765

.857

32

5

-2.256

.033 *

8

32

3

-2.756

.011 *

60

12

73

6

-3.392

.002 **

24

10

33

4

-3.274

.003 **

TLC-E a

b

Fig Lge (36 )

Note. TLC-E = Test of Language Competence-Expanded Edition. ASD = Autism spectrum disorder. TONI-2 = Test of Nonverbal Intelligence-2nd Edition. Amb Sent = Ambiguous Sentences. List Comp = Listening Comprehension: Making Inferences. Oral Exp = Oral Expression: Recreating Sentences. Fig Lge = Figurative Language. Yrs, mths = Years, months. a

= Raw performance scores on the TLC-E were used for the analysis. b = Maximum

possible score. * = p ≤ .05. ** = p ≤ .01.

Descriptively, the standard deviations for all four TLC-E subtests are greater for the ASD group compared to the control group (refer to Table 7.1), suggesting greater dispersion of scores in the ASD data set. Figures 7.1 - 7.4 display the dispersion of scores of the ASD and control adult groups’ performance on the four TLC-E subtests. There were no outliers from the ASD group on any measure. One outlier from the control group was noted on Ambiguous Sentences. Of the two groups, there is greater dispersion of

167 scores for the ASD group on Ambiguous Sentences, Listening Comprehension: Making Inferences, Oral Expression: Recreating Sentences, and Figurative Language, and on all but Oral Expression: Recreating Sentences, the middle 50 % extends further for the ASD group than the control group.

16 14

12

Amb Sent

10

8 6

4

13

2

N=

17

13

ASD GROUP

Figure 7.1.

Control


and control groups’ performance on Ambiguous Sentences.

168

14

12

List Comp

10

8

6 4

2

N=

17

ASD

Figure 7.2.

13

Control GROUP


and control groups’ performance on Listening Comprehension: Making Inferences.

169

16 14

12

Oral Exp

10

8 6

4 2

N=

17

ASD

Figure 7.3.

13

Control GROUP


and control groups’ performance on Oral Expression: Recreating Sentences.

170

16 14

12

Fig Lge

10

8 6

4 2

N=

17

13

ASD GROUP

Figure 7.4.

Control


and control groups’ performance on Figurative Language.

7.5.2

Clusters of Adults with ASD Based on Performance on the TLC-E

An agglomerative hierarchical cluster analysis was undertaken with the adults with ASD. Current performance on the four TLC-E subtests was used to determine an optimal classification. The hierarchical clustering of the ASD adults is shown in Figure 7.5.

171

Figure 7.5.

Agglomerative hierarchical cluster analysis of the 17 ASD adults

based on performance on the four subtests from the Test of Language Competence-Expanded Edition.

The analysis identified three subgroups of adults with ASD. These subgroups are descriptive only. Membership of the subgroups is shown in Table 7.2. Scores are given as within standard deviations from the control group means. Performance scores within each subgroup support the partitioning decision.

172

Table 7.2

Subgroups of ASD adults based on performance on the TLC-E

Subgroup

ID

Amb Sent

List Comp

Oral Exp

Fig Lge

1 (n = 8)

3

+1.0

+1.0

-2.0

-1.0

M age: 38 yrs (SD: 15)

5

+1.0

+1.0

-1.5

+0.5

M TONI-2: 100 (SD: 10)

1

+0.5

+1.0

-2.0

-2.0

13

+1.5

0.0

+0.5

+1.0

14

+1.0

-1.0

-1.0

+0.5

8

+1.0

0.0

+1.0

-0.5

7

-1.0

0.0

-0.5

-2.5

15

+0.5

+1.0

0.0

-3.5

2 (n = 4)

9

-0.5

-5.5

-3.0

-1.0

M age: 33 (SD: 10)

10

-1.5

-4.5

-1.5

-1.0

M TONI-2: 92 (SD: 9)

16

-1.5

-2.0

-2.0

-1.5

11

-4.0

-1.5

-1.5

+0.5 Table continues


ID

Amb Sent

List Comp

Oral Exp

Fig Lge

3 (n = 5)

4

-5.0

-4.0

-5.5

-6.5

M age: 32 (SD: 9)

17

-4.0

-4.5

-6.0

-5.5

M TONI-2: 77 (SD: 9)

2

-5.0

-5.5

-5.0

-4.5

6

-5.5

-4.5

-4.0

-4.5

12

-5.5

-7.0

-3.0

-7.0

Note. TLC-E = Test of Language Competence-Expanded Edition. ASD = Autism spectrum disorder. TONI-2 = Test of Nonverbal Intelligence-2nd Edition. Amb Sent = Ambiguous Sentences. List Comp = Listening Comprehension: Making Inferences. Oral Expression = Oral Expression: Recreating Sentences. Fig Lge = Figurative Language. Yrs = Years. Scores are entered as standard deviations above or below the control group’s subtest mean. + = Above the control group mean. - = Below the control group mean. 0.0 = No deviation from the control group mean.

173

174

Subgroup 1’s performance was characterised by generally average performance on Ambiguous Sentences and Listening Comprehension: Making Inferences, and less skilled performance on Oral Expression: Recreating Sentences and Figurative Language. Subgroup 2’s performance was strongest on Figurative Language (mild difficulty to average performance), and moderate-severe difficulties on the three remaining subtests. Subgroup 3 presented with severe difficulties spread across all four TLC-E subtests. The Kruskal-Wallis test indicated there was no significant difference between the subgroups on age (χ 2 = 1.088, p = .580). There was a significant difference between the subgroups on the TONI-2 (χ 2 = 8.949, p = .011). Post-hoc Mann-Whitney U tests determined that there was a significant difference in the TONI-2 means between subgroups 1 and 3 (Z = .000, p = .003), but no significant differences between subgroups 1 and 2 (Z = 10.500, p = .349) or subgroups 2 and 3 (Z = 3.000, p = .085). 7.6

Discussion

The aim of the present study was to examine metalinguistic language skills in adults with a diagnosis of ASD so as to provide findings that could be applied in clinical practice. The findings suggest that performance on tasks requiring metalinguistic skill was compromised in the group of adults with a diagnosis of ASD investigated in the study compared to the controls. Further, the language skills within the ASD group were not homogeneous, and subgroups within ASD could be described based on their metalinguistic competence. 7.6.1

Metalinguistic Abilities in Adults with ASD

The findings from chapter 6 of this thesis suggest that the language difficulties experienced by adults with ASD may be more extensive than previously acknowledged (see e.g., Howlin, 2003; Rumsey & Hamburger, 1988). The results of the current study extend further the clinical appreciation of the language skills associated with the disorder in adulthood, as the adults with ASD investigated in the current study experienced difficulties on a range of metalinguistic tasks relative to their peers. The ASD group of

175 participants presented with difficulties with resolving ambiguity, understanding inferential and figurative language, and producing sentence constructions that are linguistically and pragmatically appropriate to the constraints of a communicative situation. Greater heterogeneity of skill within the ASD group than in the control group was evident on all four tasks assessed in the study. It is possible that the inclusion of a range of diagnoses, such AS and HFA, may account for the non-homogeneity, given that the debate surrounding the external validation of AS from HFA is yet to be resolved (see e.g., Klin & Volkmar, 2003; Miller & Ozonoff, 1997, 2000). It is also possible that cognitive skill may account for the heterogeneity observed (see e.g., Szatmari et al., 1995). Either way, the heterogeneity suggests different individuals may present with differing support needs. If this is the case, the theoretical framework of combining nondelineated ASD participants into a single experimental group may limit the utility of the findings to clinical practice. 7.6.2

Heterogeneity of Metalinguistic Skills in Adults with ASD

The findings of the present study suggest adults with ASD exhibit a range of metalinguistic abilities as assessed by the four subtests of the TLC-E (Wiig & Secord, 1989). Subgroups could be identified in the ASD participant group, where ability ranged from above average skills through to moderate-severe impairments. The adults with ASD most proficient on metalinguistic tasks (subgroup 1) were generally able to resolve ambiguity and infer unstated information. They exhibited some linguistic flexibility by including all three words into a context-dependent sentence as required by the Oral Expression: Recreating Sentences subtest. However, violations of syntactic rules detracted from the scores for some of the subgroup participants. For instance, one participant produced Here at this bakery you will find its products fresh nor will you have to wait too long (target words are in bold). The most competent adults with ASD were generally able to provide their own interpretation to figurative speech. Performance by adults in the least competent subgroup (subgroup 3) was marked by an inability to recognise ambiguity. Most of the adults in this subgroup gave either one response or elected to pass on the Ambiguous Sentence items. When presented with “And

176 then the man wiped the glasses carefully”, one adult responded with dirty glasses and then wiped them. Generally, participants from subgroup 3 were inconsistent in selecting correct responses in the Listening Comprehension: Making Inferences subtest. In general, responses ranged from selecting two incorrect answers to selecting one of the two correctly inferred responses. Participants from this least competent subgroup exhibited difficulty with the linguistic and pragmatic constraints of the Oral Expression: Recreating Sentences tasks. For instance, one adult from the least competent group produced After the people went to the movies they bought their ticket before rather than later. The adults in this subgroup were, generally, not able to provide their own meaning for figurative language, nor were they able to derive meaning from the context provided. Most selections were literal interpretations of the target. For example, when given the situation “Two students moving to a new town” and the sentence “there is rough sailing ahead for us”, one adult provided in Venice for their own interpretation, and chose the waves are going to make it hard to sail when given four possible responses. In contrast to the finding from the preceding study where nonverbal cognitive skill was not a significant factor in subgroup membership (see section 6.6.2), nonverbal cognitive skill was significant in the subgroup formation in the present study. Nonverbal cognitive skill was a factor in subgroup formation in the study of the children with ASD based on the same metalinguistic tasks (refer to section 3.6.3). The findings from earlier studies examining a range of language and communication skills reported in this thesis have inconsistently supported the significance of nonverbal cognition to language competence within ASD. Studies investigating further the relationship between cognition and language competence are needed. 7.7

Conclusion

The results from the current study extend the clinical appreciation of the language difficulties associated with ASD in adulthood. The findings suggest that the language difficulties experienced by adults with the disorder may extend beyond gross language deficits such as receptive and expressive semantic knowledge as defined by previous studies (e.g., Howlin, 2003). Moreover, the findings suggest that proficiency on complex

177 high-level language tasks within ASD may vary, with skill level in the group of adults involved in the study ranging from above-average performance through to severe difficulty. As discussed in chapter 6, a history of developmental language disorder may have a detrimental effect on the development of social relationships and the gaining of longterm employment in adulthood (Clegg et al., 2005). Reduced psychosocial and vocational outcomes for adults with ASD have been described (Howlin et al., 2005), and the need for ASD-specific vocational services has been recognised (see e.g., Mawhood & Howlin, 1999). The high-level language tasks assessed in the current study, such as ambiguity at the word or sentence level, and understanding inferential and figurative language could be encountered on a day to day basis, and may contribute to reduced psychosocial and vocational success in adults with the disorder. Seltzer et al. (2003) have argued that appropriate support services are needed for adults with ASD diagnoses. A prerequisite for the provision of appropriate support services is an accurate clinical appreciation of the diversity of language skills in ASD. The theoretical approach of viewing ASD as a dimensional disorder and subsequently investigating subgroups may provide a clinical understanding of individual strengths and weaknesses, which can be incorporated into individualised support plans. Chapters 6 and 7 of this thesis focus on linguistic, pragmatic, and abstract language skills of the adults with a diagnosis of ASD. Following Manjiviona's (2003) suggestion, the following chapter reports on their social problem solving skills.

178

8

Social Problem Solving in Adults with Autism Spectrum Disorder

8.1

Introduction

Difficulties with social problem solving may contribute to the social and emotional deficits associated with the autistic presentation (Solomon et al., 2004). Social enhancement programmes have, therefore, included activities aimed at developing social problem solving abilities in children with ASD (Solomon et al.). Solomon et al. undertook a control group study of a social enhancement intervention programme for children with HFA, AS, and pervasive developmental disorder-not otherwise specified. The intervention programme was designed to target emotional recognition in self and others, ToM (including perspective taking) and executive functions with a particular emphasis on individual and group problem solving. Pre- and post-intervention measures indicated that high functioning children with ASD could be taught real life problem solving skills within the context of social enhancement programmes. Few studies to date have examined social problem solving in the adult ASD population. Nevertheless, a number of service providers catering to the needs of adults with ASD have developed programmes which include social skill enhancement (e.g., Alpha Autism Inc., 2007; NAS, 2007). An accurate clinical profile of the social problem solving skills in adults with the disorder would, therefore, be necessary to determine if such social skill enhancement programmes should include a focus on social problem solving. 8.2

Verbal and Social Problem Solving in Adults with ASD

Previous research findings have suggested that verbal and/or social problem solving may be sub-optimal in adults with ASD (e.g., Minshew, Goldstein, Muenz, & Payton, 1992; Minshew, Siegel, Goldstein, & Weldy, 1994; Rumsey & Hamburger, 1988). Rumsey and

179 Hamburger (1988), for instance, administered a comprehensive neuropsychological test battery to a group of adult men with a childhood diagnosis of Infantile Autism as defined by DSM-III criteria (APA, 1980). All men had verbal and performance IQs at or above 80 as measured by the Wechsler Adult Intelligence Scale (Wechsler, 1955). As reported in section 6.1 of this thesis, Rumsey and Hamburger’s participants presented with mild linguistic difficulties. The researchers aimed to investigate whether formal measures of problem solving and reasoning would document deficits in the group, and whether the problem solving difficulties would be pervasive or limited to verbal measures only. Rumsey and Hamburger (1988) administered the Verbal Absurdities task and the Problem Situations task from the Stanford-Binet Intelligence Scale (Terman & Merrill, 1960) to assess complex verbal problem solving. The autistic group performed significantly below a control group matched for age and education on both verbal problem solving measures. This finding is consistent with the proposal that language skills influence problem solving (e.g., Baldo et al., 2005; Liss et al., 2001; Solomon et al., 2004). Moreover, the autistic group was impaired on the nonverbal problem solving tasks, suggesting that pervasive problem solving deficits are associated with the disorder. An extensive neuropsychological assessment of nonmentally retarded individuals with AD was likewise undertaken by Minshew et al. (1992). Participants in Minshew et al.’s study were aged between 15 – 40 years and met DSM-III-R (APA, 1987) criteria for AD, with the diagnoses validated by the Autism Diagnostic Interview [ADI] (LeCouteur et al., 1989) and the Autism Diagnostic Observation Schedule [ADOS] (Lord et al., 1989). The Test of Language Competence (Wiig & Secord, 1985) was administered to examine the use of language in problem solving contexts. The autistic group of adults presented with significant linguistic problem solving deficits relative to a control group matched on age, full scale IQ, gender, and race. In a later study, Minshew et al. (1994) examined verbal problem solving in a group of high functioning individuals (full scale and verbal IQ over 70) with diagnoses of autism validated by the ADI (LeCouteur et al., 1989) and the ADOS (Lord et al., 1989). Using the Twenty Questions Procedure (Mosher & Hornsby, 1966; Olver & Hornsby,

180 1966), Minshew and colleagues examined the ability of the autistic participants to conceptualise information in verbal problem solving. The researchers described similar levels of efficiency (time taken and the number of questions used) between the autism group and a typically developing control group matched on age, sex, years of education, race, and socio-economic status. In terms of using effective strategies to achieve the solutions, the control group was more proficient than the autism group. 8.3


The studies detailed above (Minshew et al., 1992; Minshew et al., 1994; Rumsey & Hamburger, 1988) provide consistent findings of difficulties with verbal problem solving in autistic adults with average intelligence. There are, however, limitations to the studies which, if addressed, could extend the current understanding of problem solving in ASD. Firstly, the three studies focused on verbal problem solving skills. No known study has focused on real-life problems that may be encountered in social situations, such as the studies investigating real-life problem solving, or social problem solving, in children and adolescents with AS undertaken by Channon et al. (2001) and Griswold et al. (2002) respectively. A clinical appreciation of the social problem solving capabilities in adults with ASD is necessary, particularly if, as suggested by Charman (2004), poor skill at resolving problem situations may cause significant impairment in every day life. Secondly, due to the methodological requirements of the studies, no matching for verbal ability was undertaken by Rumsey and Hamburger (1988) or Minshew and coresearchers (1992; 1994) in their earlier or later study. Although Rumsey and Hamburger described the problem solving difficulties in their group of high functioning autistic men as pervasive, it would be clinically beneficial to determine whether verbal or social problem solving deficits are evident when the effect of language ability is removed. This would have clinical relevance, as at this point in time, it is not known if the difficulties are specific to problem solving or whether the difficulties are language-based. This issue needs to be resolved before any intervention, such as the inclusion of social problem solving in social enhancement programmes, can be considered.

181 Thirdly, Rumsey and Hamburger (1988) documented a high degree of heterogeneity of language skills within their autistic sample. Moreover, the standard deviations of Rumsey and Hamburger’s autistic group were far greater than the standard deviations of the control group on the Verbal Absurdities and Problem Situations tasks (Terman & Merrill, 1960), suggesting that verbal problem solving abilities in ASD may likewise be marked by a high degree of heterogeneity. In contrast, the results of Minshew et al.’s (1992) study suggested greater homogeneity in the autistic presentation, as the standard deviations of Minshew et al.’s autistic group on two of the four Test of Language Competence subtests (Wiig & Secord, 1985) were smaller than the standard deviations of the control group. Further investigation is required to clarify nature of the ASD presentation on social problem solving, as the development of intervention and support programmes requires an accurate clinical understanding of the difficulties associated with the disorder. 8.4


The present study aimed to redress the methodological limitations of studies to date by investigating social problem solving abilities in a group of adults with a diagnosis of ASD. Due to the nature of the investigations undertaken in chapters 6 and 7, the ASD participants could not be matched to control participants on verbal ability. Nevertheless, to extend the findings from previous studies (Minshew et al., 1992; Minshew et al., 1994; Rumsey & Hamburger, 1988), the effect of language skill was removed in the present study. Based on the theoretical framework of combined studies (see section 1.3.2), performance by the adults with a diagnosis of ASD on social problem solving tasks was compared to the performance of a control group with non-significant developmental histories. The framework underlying spectrum studies (as outlined in section 1.3.3) could not be applied due to the limited number of measures available for the hierarchical cluster analysis. Heterogeneity within the ASD group of participants was examined through score dispersal only. No known study has removed the effect of language in investigations of social problem solving in adults with ASD, therefore, no prediction was made regarding the

182 performance on tasks of social problem solving by the ASD group compared to the control group. As heterogeneity within the autistic presentation has been noted on tasks of verbal problem solving (Rumsey & Hamburger, 1988), it was predicted that the ASD group investigated in the present study would exhibit non-homogeneity in social problem solving abilities. 8.5

Method

8.5.1

Participants

Recruitment details of the 17 adults with a diagnosis of ASD and the 13 control participants are described in section 6.4.1. The ASD participant demographics are displayed in Table 6.1. 8.5.2

Measures

All participants were assessed using the TOPS-A (Bowers et al., 1991), the TONI-2 (Brown et al., 1990), and the WAB (Kertesz, 1982). The TOPS-A is described in section 5.5.2. The purpose of the TONI-2 to the current investigation is discussed in section 2.4.2. The WAB is discussed in detail in section 6.4.2. 8.5.3

Procedure

The TOPS-A (Bowers et al., 1991) is not standardised for the adult population, but was chosen for the present study as there is no equivalent test standardised for adults. Standard Scores, corrected for age, are provided for a maximum age of 17 years 11 months only. Therefore, raw scores from the TOPS-A were used for the statistical analyses. The ASD group of participants had significantly less proficient linguistic skills relative to the control group (t = -4.130; p = .001) as assessed by the WAB AQ (Kertesz, 1982) (see section 6.5.1). There was no significant difference between the two groups on nonverbal cognitive skill (t = -1.765; p =.857) as assessed by the TONI-2 (Brown et al., 1990) (see section 6.5.1), and no difference in age (t = 0.040; p =.968). As the TOPS-A

183 assesses language-based critical thinking (Bowers et al., 1991), correlations between verbal skill and problem solving skills were investigated. An analysis of covariance (ANCOVA), entering the WAB AQ score as a covariant and the TOPS-A as the dependent variable, was undertaken to remove the effect of verbal skills. 8.6

Results

Means, standard deviations, and ranges of the ASD and control groups’ performance scores on the measures of verbal and problem solving skills are displayed in Table 8.1. There was a highly significant correlation between performance on the WAB AQ and performance on the TOPS-A (r = .790; p = .001) for the ASD group, but not for the control group (r = .352; p = .238).

Table 8.1 Means, standard deviations, and ranges of performance scores from the adult participants on the verbal and problem solving measures

Variable

ASD

Control

(N = 17)

(N = 13)

M

SD

Range

M

SD

Range

Mean diff.

WAB AQ

95.90

3.40

89.3-99.7

99.50

1.10

96.1-100

3.6

TOPS-A (50 a)

26

9

11-39

42

4

35-49

16

Note. ASD = Autism spectrum disorder. WAB AQ = Western Aphasia Battery Aphasia Quotient. .TOPS-A = Test of Problem Solving Adolescent. Mean diff = Differences in the means of the two groups. a

= Maximum possible raw score.

ANCOVA identified a highly significant group difference on the TOPS-A (F (1, 28) = 11.537, p = .002) when the effect of language skill was removed. Of the two groups, the ASD group was significantly less skilled on social problem solving.

184 Descriptively, the ASD group’s standard deviation on the TOPS-A is larger than the control group’s (refer to Table 8.1), suggesting greater score dispersal for the ASD group. Figure 8.1 displays the spread of scores from the ASD and control adult groups’ performance on the TOPS-A. Outliers were noted for the control group. The length of the box plot for the ASD group far exceeds the length of the box plot for the control group, and the middle 50 % of ASD scores extends further than the control group’s scores, both indicating greater score dispersal for the ASD group.

60

12 50

40

TOPS-A

7 30

20

10 0

N=

17

ASD

Figure 8.1.

13

Control GROUP


and control groups’ performance on the Test of Problem Solving-Adolescent.

185 8.7

Discussion

Using one of the three theoretical approaches outlined in chapter 1 and a descriptive analysis of score dispersal, the current research aimed to extend the clinical understanding of social problem solving skills in a group of adults with a diagnosis of ASD. The findings suggest that the group of adults with ASD investigated in the study had significantly poorer social problem solving skills relative to the control group when language skill was removed from the analysis. Of the two groups, the ASD group displayed greater heterogeneity in social problem solving capabilities. 8.7.1

Social Problem Solving in Adults with ASD

In the present study, social problem solving skill was investigated in a group of adults with ASD through an assessment of language-based critical thinking (Bowers et al., 1991). It has been proposed that verbal skills may strongly influence reasoning and problem solving (Baldo et al., 2005; Liss et al., 2001), but there was no significant correlation between linguistic skills and problem solving for the control group in the present study. This may, however, have been due to a ceiling effect on the WAB AQ (Kertesz, 1982). In contrast, and offering support for Baldo et al.’s contention that problem solving is influenced by verbal skill, there was a highly significant correlation between linguistic skill and social problem solving for the group of ASD participants. Consequently, linguistic skill as a variable was removed from the analysis of social problem solving. When the effect of verbal differences between the ASD group and the control group was taken into account, the ASD group presented with significantly impaired social problem solving skills. Hence, even if the language skills of the ASD group were commensurate with the control group, the ASD group would still present with sub-optimal social problem solving skills. 8.7.2

Heterogeneity of Social Problem Solving Skills in Adults with ASD

The limitation of using Standard Scores to examine the spread of scores in the study of social problem solving in the children with ASD has previously been identified (see

186 section 5.7.1). In the present study, this limitation was avoided by using raw scores which did not mask the actual dispersal of scores. Consistent with Rumsey and Hamburger’s (1988) study, the present research identified greater heterogeneity within the ASD group of participants than in the control group. The examination of the heterogeneity of skill of the two participant groups through the use of box plots was based on the spread of TOPSA (Bowers et al., 1991) performance scores alone. Language skill was not controlled in the descriptive analysis of the spread of scores. The finding of heterogeneity in social problem solving in the present study is unsurprising, given the significant relationship between language skill and performance on an assessment of language-based critical thinking, combined with the heterogeneity of language skills within the ASD group of participants noted in previous chapters of this thesis (see sections 6.6.2 and 7.6.2). Although statistical analysis of the heterogeneity was not undertaken, qualitative differences within the ASD group could be described. A comparison of the quality of responses provided by the ASD participants suggests that the individuals investigated in the present study may have differing support needs in social problem solving situations. The following presents an example of the qualitative differences in performance between ASD participant 1 and ASD participant 2. Participant 1 was one of the higher ASD TOPS-A scorers while participant 2 was one of the lower scorers from the group. “Holiday Choices”, one of the stories presented in the TOPS-A (Bowers et al., 1991) is as follows: “Marnie’s parents are divorced, and she’s an only child. She lives with her mother. They all live in the same town, and Marnie visits her father one weekend each month. Her parents don’t get along. Marnie’s father wants her to go on a skiing trip with him during her holiday vacation. That’s the only time he can take the ski trip. Marnie’s mother wants her home for the holidays because she’s having a family celebration”. In response to the question “What’s Marnie’s problem?”, ASD participant 1 responded both parents want her to go with them. ASD participant 2 responded Marnie doesn’t like her parents much. The answer to the question is not stated in the story, and inference is required to answer the question. Participant 1 correctly inferred the problem. There is nothing in the story or the context of the story, however, to support participant

187 2’s assumption regarding Marnie’s opinion of her parents. The next question is “Who do you think should make the decision about how Marnie spends her vacation? Why?”. Participant 1 thought Marnie should make the decision, because if she makes the decision her parents won’t have to fight about who has the most control over Marnie. Participant 2 also thought Marnie should make the decision, because her parents might not like where she goes. The final question is “Now suppose Marnie’s parents decide to let Marnie make her own choice. What compromise could Marnie suggest to her parents about the holiday?”. Participant 1 suggested the ski trip could be at any time, but the family celebration can’t, while participant 2 suggested the mother and father can go on the ski trip. Participant 2’s response can not be viewed as an appropriate solution that would please the three individuals concerned for a number of reasons, the most obvious being that the parents do not get along. The quality of participant 2’s responses suggests that this participant may encounter considerably more breakdowns or misunderstandings in social interactions than participant 1. Diagnoses of AS or HFA were combined into one experimental group for the current analysis. As research studies investigating the validation of AS from HFA continue (see e.g., Klin et al., 2005; Macintosh & Dissanayake, 2004), it is, therefore, not possible to interpret if combining the two diagnostic categories into a non-delineated experimental group influenced the findings. 8.7.3

Clinical Utility of the Findings

The overall findings of the current study suggest that the adults with ASD presented with considerable difficulties in social problem solving. Moreover, the performance by the ASD group on social problem solving tasks was more heterogeneous than that of the control group, suggesting that some adults with ASD may have far greater need of support in social problem solving situations than others on the spectrum. If, as argued by Solomon et al. (2004), impaired social problem solving contributes negatively to the social and emotional functioning of individuals with ASD, the findings could be interpreted as having clinical utility in terms of offering support for the inclusion of a focus on problem solving behaviours in social enhancement programmes.

188 As presented in the study investigating social problem solving in children with ASD (see section 5.7.2), however, numerous skills may contribute to effective social problem solving, and could account for both the significantly poorer social problem solving abilities of the ASD participants as a group and the heterogeneity noted within the group. Language skill may, for instance, influence social problem solving (Liss et al., 2001). Language skill was controlled as a covariant in the between-group analysis, but not in the descriptive analysis of the spread of scores within the two participant groups. Language competence may, therefore, be a contributing factor in the heterogeneity noted in the ASD group. The ToM deficits described in the ASD population (e.g., Hill, 2004) may result in difficulty inferring the mental states of others (Rutherford, Baron-Cohen, & Wheelwright, 2002). Aspects of compromised performance by the ASD group could reflect ToM deficits. Question 27, for instance, (“Jason has a choice now. What do you think he’ll do? Why”?), requires the participant to infer the mental state of another, or in other words, apply ToM. Individuals with ASD may have weakened central coherence (e.g., Baron-Cohen, 1988), which could influence performance on the tasks from the TOPS-A (Bowers et al., 1991). As discussed in section 5.7.2, central coherence is the ability to draw together information and to establish causal links between linguistic elements, and enables a person to integrate information to derive the full meaning or gestalt (see e.g., Jolliffe & Baron-Cohen, 2000; Noens & van Berckelaer-Onnes, 2005). Jolliffe and Baron-Cohen described weakened central coherence in adults with ASD, and suggested the inability to fully appreciate situations may be due to a failure to interpret, within the wider context, the given information. Numerous aspects of the TOPS-A stories need to be integrated to gain a full understanding of the story. For instance, a full appreciation of the situation facing Marnie in “Holiday Choices” (presented above) is needed to adequately answer the questions relating to the story. ASD participant 2’s responses, as detailed above, may reflect weakened central coherence, as he missed the gist of the story, and hence, provided inappropriate responses to the questions. Executive function skills, such as goal setting, devising strategies, planning,

189 cognitive flexibility, self-monitoring, and judgment (Bradshaw, 2001; Hill, 2004) have been described as deficient in disorders on the autism spectrum (Fletcher-Watson, Leekam, Turner, & Moxon, 2006; Landa & Goldberg, 2005; Russell & Jarrold, 1998). It is possible that the problem solving difficulties evident in the ASD group reflect underlying executive function deficits (Charman, 2004). Apart from language skill, the influence of the above-listed variables on problem solving was not accounted for in the present study. The clinical utility of the findings is, therefore, limited at this point in time as it is not possible, from the study’s findings, to isolate social problem solving per se as the sole contributor to the impaired performance by the ASD group. In the same vein, it may not be possible to determine how best to develop individual and group problem solving strategies for inclusion into social enhancement intervention programmes, such as the programme devised by Solomon et al. (2004), when numerous variables may influence an individual’s ability to solve problems in social situations. 8.8

Specific Limitations and Future Research Directions

The present study has a number of specific limitations, most of which are presented in the study examining social problem solving in children with ASD (e.g., not controlling for variables such ToM, central coherence, and executive function skill, and the lack of interrater reliability on scoring decisions). Additionally, the ASD participants’ diagnoses included both AS and HFA, and unlike the child studies, the validation approach was not applied to determine if AS and HFA could be differentiated. There is ongoing research into whether AS and HFA are similar autistic disorders, and executive skill has been one measure of interest in differentiating the two presentations (see e.g., Ozonoff, South, & Miller, 2000; Rinehart, Bradshaw, Brereton, et al., 2002; Rinehart et al., 2001; Rinehart, Bradshaw, Tonge, Brereton, & Bellgrove, 2002; Verte, Geurts, Roeyers, Oosterlaan, et al., 2006). The inclusion of both diagnoses into one experimental group may have influenced the findings, particularly if executive function skills, which may reduce problem solving capabilities (Charman, 2004), differentiate AS from HFA.

190 Replication of the study with a larger number of participants is needed. Studies where participants with ASD are matched to controls on verbal IQ would be of interest, particularly in terms of examining further the nature of the heterogeneity within the ASD sample. Additionally, studies investigating the interplay between problem solving and cognitive attributes such as ToM and central coherence, and the relationship of language to problem solving would be of interest. Given 67 % of the adults with ASD investigated in the present study who were in a position to work full-time were unemployed, future research should investigate the relationship between social problem solving skills and vocational outcomes. 8.9

Conclusion

The ASD group of participants investigated in the current study presented with less competent social problem solving skills than the control group when the effect of language skill was removed. A high degree of heterogeneity of skills within the ASD group was noted, but not further assessed through within-group analysis. Nonetheless, the findings suggest that factors other than language skill must contribute to competence in social problem solving in the adults with ASD. The studies in this thesis presented thus far serve to extend the clinical understanding of the language and communication difficulties experienced by children (chapters 2 - 5) and adults (chapters 6 - 8) with a diagnosis of ASD. There is, however, one group within the ASD population of which, clinically, little is known. To date, no known research has investigated the language and communication skills of females with the diagnosis. The final study of this thesis, presented in the next chapter, seeks to redress this gap in research.

191

9

Females and Males with Autism Spectrum Disorder: A Comparison Study of Language and Social Pragmatic Skills 6

9.1

Introduction

Wing (1989) introduced the concept of an autistic continuum or spectrum as a means of acknowledging all individuals presenting with the triad of autistic characteristics. The preceding studies in this thesis report on communicative difficulties that may be experienced by children and adults with the diagnosis. The findings of the preceding chapters have contributed valuable clinical insights into the strengths and weaknesses in the language and communication skills of the children and adults with ASD who participated in the research project. Yet, the preceding studies have not investigated the clinical presentation of the language and communication skills of all individuals with ASD represented in the study, as the female presentation of ASD has not been considered as a separate variable thus far in this thesis. ASD can occur in both males and females (Baron-Cohen, 2002; Gillberg, 1989; Howlin & Asgharian, 1999; Kadesjo et al., 1999; Wing, 1981). It has been postulated, however, that the autism phenotype may differ between females and males (Kopp & Gillberg, 1992), that the characteristics suggestive of autistic-like conditions emerge at an earlier age in males than in females (Gillberg & Rastam, 1992), or that females may have less obvious language and social skill deficits than males with the disorder (Frith, 2003;

6

“Females with autistic features: Are language and social skills barriers to diagnosis?”, by F. M. Lewis,

G. C. Woodyatt, and B. E. Murdoch, 2006, 13th Learning Conference, Jamaica, was adapted from this chapter.

192 Kopp & Gillberg, 1992). If the presentation of ASD differs in males and females, the DSM-IV (APA, 1994) criteria for AS and AD fail to take this into account, as there are no gender-related guidelines in these criteria. To date, there is a dearth of information on how females present on the spectrum of autistic disorders. While many studies have included females in their ASD participant groups (e.g., Emerich et al., 2003; Nyden, Billstedt, Hjelmquist, & Gillberg, 2001; Prior et al., 1998; Seltzer et al., 2003; Sturm, Fernell, & Gillberg, 2004), no known studies have undertaken within-group analyses to determine whether the female presentation is similar to the presentation in males. In response to the lack of understanding of the presentation of ASD in females, Kaland et al. (2002) and Nyden et al. have cautioned against generalising research findings to females. A possible difference in language skills between males and females with ASD has been suggested by Kopp and Gillberg (1992). They proposed the existence of a female phenotype of ASD, where girls have stronger language skills and use more imitative skills in social interactions. On the same theme, Frith (2003) hypothesised that females may have relatively good language skills, be compliant in educational settings, and may adopt compensatory learning styles. This combination of skills may obscure the central deficits to a large extent, and may result in greater handicap for females than males with ASD (Kopp & Gillberg), as it could reduce the apparent need for intervention and support. 9.2


Kopp and Gillberg (1992) and Frith (2003) have both proposed that females with ASD may present with better language and social skills, yet no study has, to date, examined the proposal. The ASD participants involved in the preceding studies have included a number of females (four in the child studies and nine in the adult studies). Given the almost equal number of females and males in the adult studies (9 females: 8 males), a pilot study examining the language and social skills of the nine female adults is possible. If females and males with ASD have differing communicative profiles, they may likewise present

193 with differing intervention needs. The study will, therefore, provide preliminary clinical findings of the communicative needs of the female adults with ASD. 9.3

Aim and Hypothesis of the Present Study

The aim of the current study was to compare the language and pragmatic social skills of females with a diagnosis of ASD with the skills of males with the diagnosis. To provide comparative data, the performance by the females and the males with the diagnosis was compared to females and males with a typical developmental history. Based on previous untested hypotheses (e.g., Frith, 2003; Kopp & Gillberg, 1992), it was hypothesised that females with ASD would present with better language and social skills when compared to males with ASD. 9.4

Method

9.4.1

Participants

Recruitment details of the 17 adult participants with a diagnosis of ASD and the 13 control participants are described in section 6.4.1. Nine of the ASD participants were female (M age: 32 years; SD: 9 years; range: 18 years – 48 years; M years of education: 12 years 2 months; range: 8 years – 15 years) and eight were male (M age: 38 years; SD: 15 years; range: 27 years – 67 years: M years of education: 12 years 7 months; range: 10 years – 15 years). Table 9.1 displays the demographic details of the ASD participants grouped according to gender.

194

Table 9.1

Female and male ASD adult participant

demographics ID

Sex

Age (years)

Ed level

Diag

(Fe: M = 9: 8) 1

Fe

18

Sec

HFA

2

Fe

30

Post sec

AS

3

Fe

42

Sec

Autism

4

Fe

31

Tert

AS

5

Fe

22

Post sec

AS

6

Fe

26

Sec

AS/ASD

7

Fe

38

Tert

AS

8

Fe

32

Sec

AS

9

Fe

48

Prim

AS

10

M

29

Sec

AS

11

M

27

Tert

AS

12

M

28

Sec

HFA

13

M

48

Tert

AS

14

M

29

Post sec

AS

15

M

30

Tert

AS

16

M

67

Post sec

AS

17

M

48

Post sec

AS

Note. ASD = Autism spectrum disorder. AS = Asperger syndrome. HFA = High functioning autism. Diag = Independently determined diagnosis. Ed level = Level of education. Prim = Primary. Sec = Secondary. Post sec = Post-secondary vocational trade. Tert = Tertiary. Fe = Female. M = Male

195

Seven females (M age: 30 years; SD: 9 years; range: 18 years – 44 years; M years of education: 13 years 1 month; range: 12 years – 15 years) and six males (M age: 40 years; SD: 17 years; range: 24 years – 65 years; M years of education: 13 years; range: 11 years – 15 years) formed the control group. Statistical analysis indicated there were no significant differences between the four groups on age (χ 2 = 1.794; p = .616) or nonverbal cognitive skills (χ 2 = 2.943; p = .400) as assessed by the TONI-2 (Brown et al., 1990). 9.4.2

Measures

A comprehensive test battery was devised to assess language abilities and pragmatic social skills. The language test battery consisted of the WAB (Kertesz, 1982), the RHLB (Bryan, 1989), the TOPS-A (Bowers et al., 1991), the TLC-E (Wiig & Secord, 1989), the RSPCS-SR, and the RSPCS-SO (Burns et al., 1985). These assessments are described in earlier studies presented in this thesis. Details of these assessments are presented in Appendix J. 9.4.3

Procedure

The WAB AQ (Kertesz, 1982) was used as a measure of overall linguistic ability. The WAB subtests used to obtain the WAB AQ are described in section 6.4.2. As outlined in section 6.4.3, two RSPCS (Burns et al., 1985) measures were obtained: a self-rating (RSPCS-SR) and a rating by a person familiar with the participant (RSPCS-SO). 9.5

Results

The means and standard deviations of all measures for the four participant groups are displayed in Table 9.2. For significance at the 0.05 level, p must be equal to or less than 0.003, as determined by the Bonferroni adjustment to correct for the multiplicity of tests (Price, 2000).

196

Table 9.2

Descriptive statistics of the four adult comparative groups ASD Females

Control Females

ASD Males

Control Males

Kruskal-Wallis Test

(n = 9)

(n = 7)

(n = 8)

(n = 6)

Variable

M (SD)

M (SD)

M (SD)

M (SD)

χ2

p

WAB AQ

95.13 (3.61)

99.20 (1.49)

96.72 (3.09)

99.83 (0.27)

16.837

.001 ***

Amb Sent (39 a)

26 (11)

31 (5)

24 (13)

33 (7)

2.911

.406

List Comp (36 a)

26 (9)

30 (4)

25 (7)

33 (2)

5.548

.136

Oral Exp (78 a)

63 (12)

72 (6)

58 (11)

73 (6)

10.590

.014

Fig Lge (36 a)

24 (10)

32 (5)

24 (10)

33 (3)

7.359

.061

MP

53 (9)

61 (1)

52 (12)

61 (1)

8.463

.037

WM

53 (8)

57 (3)

54 (7)

58 (3)

2.155

.541

CIM

41 (6)

52 (5)

42 (9)

52 (3)

15.899

.001 ***

AH

47 (12)

56 (7)

46 (14)

60 (4)

6.851

.077

TLC-E

RHLB

Table continues

Table 9.2 continued ASD Females

Control Females

ASD Males

Control Males

Kruskal-Wallis Test

(n = 9)

(n = 7)

(n = 8)

(n = 6)

Variable

M (SD)

M (SD)

M (SD)

M (SD)

χ2

p

LS

59 (7)

62 (0)

55 (9)

62 (0)

7.412

.060

PES

44 (5)

56 (2)

48 (7)

56 (2)

20.238

.001 ***

TOPS-A (50 a)

26 (8)

42 (5)

27 (12)

41 (2)

19.546

.001 ***

RSPCS-SR (60 a)

34.91 (11.83) (8 b)

51.54 (9.21)

39.47 (5.39)

52.50 (5.61)

13.863

.003 **

RSPCS-SO (60 a)

39.94 (7.52)

53.29 (4.76) (6 b)

35.53 (9.28)

53.08 (7.95)

16.254

.001 ***

Note. ASD = Autism spectrum disorder. WAB AQ = Western Aphasia Battery Aphasia Quotient. TLC-E = Test of Language Competence-Expanded Edition. Amb Sent = Ambiguous Sentences. List Comp = Listening Comprehension: Making Inferences. Oral Exp = Oral Expression: Recreating Sentences. Fig Lge = Figurative Language. RHLB = Right Hemisphere Language Battery. M P = Metaphor Picture Test. W M = Written Metaphor Test. C I M = Comprehension of Inferred Meaning. A H = Appreciation of Humour. L S = Lexical Semantic Test. P E S = Production of Emphatic Stress. TOPS-A = Test of Problem Solving-Adolescent. RSPCS-SR = Rating Scale of Pragmatic Communication Skills-Self-Rating. RSPCS-SO = Rating Scale of Pragmatic Communication Skills-Rating by a Significant Other. a

= Maximum possible score. b = Missing data. The number preceding b indicates the number of cases involved in the analysis.

** = p ≤ .003 (Bonferroni adjustment). *** = p ≤ .001.

197

198 As displayed in Table 9.2 (refer to the previous pages), the Kruskal-Wallis test identified there were no significant differences between the four groups on Ambiguous Sentences, Listening Comprehension: Making Inferences, Oral Expression: Recreating Sentences, Figurative Language, Metaphor Picture Test, Written Metaphor Test, Appreciation of Humour, or Lexical Semantic Test (p ≥ .003 on all measures). Significant differences between the groups were found on the WAB AQ (p ≤ .001), Comprehension of Inferred Meaning (p ≤ .001), Production of Emphatic Stress (p ≤ .001), TOPS-A (p ≤ .001), the self-rating on the RSPCS (p ≤ .003), and on the RSPCS rated by a significant other (p ≤ .001). Table 9.3 (on the following page) displays the results of post-hoc analyses using Mann-Whitney U tests. For significance at the 0.05 level, p must be equal to or less than 0.008, as determined by the Bonferroni adjustment to correct for the multiplicity of tests (Price, 2000).

Table 9.3

Variable

WAB AQ

Language and social skill differences between ASD females, ASD males, control females, and control males ASD Female /

ASD Male /

ASD Female /

Control Female /

Control Female

Control Male

ASD Male

Control Male

Z

Z

p (Exact

p (Exact

Z

p (Exact

Z

p (Exact

Sig. 1-

Sig. 1-

Sig. 1-

Sig. 1-

tailed)

tailed)

tailed)

tailed)

-2.719

.005 **

-2.679

.005 **

-1.443

.167

-0.489

.731

CIM

-2.894

.002 **

-2.688

.005 **

-0.389

.743

-0.148

.945

PES

-3.397

.001 ***

-2.646

.008 **

-1.745

.093

-0.488

.731

TOPS-A

-3.233

.001 ***

-2.973

.001 ***

-0.531

.606

-0.574

.628

RSPCS-SR

-2.201

.029

-2.976

.001 ***

-1.052

.328

-0.215

.836

RSPCS-SO

-2.715

.005 **

-2.711

.005 **

-0.819

.423

-0.323

.818

RHLB

Note. ASD = Autism spectrum disorder. WAB AQ = Western Aphasia Battery Aphasia Quotient. RHLB = Right Hemisphere Language Battery. C I M = Comprehension of Inferred Meaning. P E S = Production of Emphatic Stress. TOPS-A = Test of Problem SolvingAdolescent. RSPCS-SR = Rating Scale of Pragmatic Communication Skills-Self Rating. RSPCS-SO = Rating Scale of Pragmatic Communication Skills-Rating by a Significant Other. ** = p ≤ .008 (Bonferroni adjustment). *** = p ≤ .001.

199

200

Table 9.4

Correlations between the Significant Other rating on the RSPCS and measures of language

performance for the four comparative adult groups Variable

ASD female

Control female

ASD male

Control male

Spearman’s rho

Spearman’s rho (p)



(p level) TONI-2

.234 (.544)

.812 (.050)

.719 (.045)

.314 (.544)

WAB AQ

.385 (.306)

.789 (.062)

.429 (.289)

.338 (.512)

Amb Sent

-.117 (.764)

-.255 (.582)

.398 (.329)

.294 (.572)

List Comp

-.437 (.240)

-.708 (.075)

.217 (.606)

-.266 (.611)

Oral Exp

-.243 (.529)

-.667 (.101)

.310 (.456)

.045 (.933)

Fig Lge

-.538 (.135)

-.841 (.018)

.335 (.417)

-.206 (.696)

MP

.149 (.702)

.399 (.434)

.272 (.515)

.393 (.441)

WM

.715 (.030)

-.105 (.843)

.464 (.247)

-.393 (.441)

CIM

.271 (.480)

.806 (.053)

.359 (.382)

.395 (.439)

TLC-E

RHLB

Table continues

Table 9.4 continued Variable

ASD female

Control female

ASD male

Control male

Spearman’s rho




(p level) AH

.542 (.131)

.585 (.222)

.218 (.604)

-.372 (.468)

LS

.550 (.125)

%%

-.421 (.298)

%%

PES

.506 (.164)

.105 (.843)

.638 (.089)

-.655 (.158)

TOPS-A

-.481 (.190)

-.288 (.531)

.214 (.610)

.500 (.313)

Note. RSPCS = Rating Scale of Pragmatic Communication Skills. ASD = Autism spectrum disorder. TONI-2 = Test of Nonverbal Intelligence-2nd Edition. WAB AQ = Western Aphasia Battery Aphasia Quotient. TLC-E = Test of Language Competence-Expanded Edition. Amb Sent = Ambiguous Sentences. List Comp = Listening Comprehension: Making Inferences. Oral Exp = Oral Expression: Recreating Sentences. Fig Lge = Figurative Language. RHLB = Right Hemisphere Language Battery. M P = Metaphor Picture Test. W M = Written Metaphor Test. C I M = Comprehension of Inferred Meaning. A H = Appreciation of Humour. L S = Lexical Semantic Test. P E S = Production of Emphatic Stress. TOPS-A = Test of Problem Solving-Adolescent. %% = Perfect correlation of 1.00 due to maximum possible score obtained.

201

202 As shown in Table 9.3 (see page 199), there were no significant differences between the control female group and the control male group on all measures (p ≥ .008 on all comparisons), and no significant differences between the ASD female group and the ASD male group (p ≥ .008 on all measures). Significant differences were identified between the female ASD group and the female control group on the WAB AQ (p ≤ .008), Comprehension of Inferred Meaning (p ≤ .008), Production of Emphatic Stress (p ≤ .001), TOPS-A (p ≤ .001), and RSPCS-SO (p ≤ .008), with the ASD female group less competent on all measures. There was no significant difference between the two female groups on RSPCS-SR (p ≥ .008). Significant differences on all measures were detected between the male ASD group and the male control group. The male ASD group performed significantly below their gender-matched control group on the WAB AQ (p ≤ .008), Comprehension of Inferred Meaning (p ≤ .008), Production of Emphatic Stress (p ≤ .008), TOPS-A (p ≤ .001), RSPCS-SR (p ≤ .001), and RSPCS-SO (p ≤ .008). To determine if there was a relationship between language competence and RSPCS-SO rating, Spearman’s correlations were undertaken for all four participant groups. Apart from the perfect correlations between the RSPCS-SO and the Lexical Semantic Test for both the control female and control male groups due to both groups obtaining the maximum score possible on the Lexical Semantic Test, there were no significant correlations between the Significant Other rating on the RSPCS or any of the language variables for any of the groups (refer to Table 9.4 on pages 200-201). 9.6

Discussion

The current study aimed to profile the language and social skills of females with ASD. The prediction that the females with ASD investigated in the study would present with better language and social skills relative to the male ASD presentation was not supported by the findings. A comprehensive language battery failed to isolate any language measure that differentiated the females with ASD from the males with the diagnosis. Additionally, there was no difference in pragmatic social skills between the females and the males with ASD.

203 9.6.1

Language Skills in Females with ASD

Previous researchers (Frith, 2003; Kopp & Gillberg, 1992) have raised the possibility that females with ASD have stronger language and social skills, or possibly less-obvious deficits in these two skill areas. It is difficult to evaluate this proposal without knowing the standard against which the language skills of females with ASD are compared. If the standard for comparison is the male ASD population, the results of the current study fail to offer support, as the females with ASD assessed in the current study presented with similar language deficits to their male counterparts. If the standard against which the language skills of females with ASD are compared is their gender-matched peers, this also fails to support the hypothesis that females may present with less-obvious language deficits. Despite there being no significant difference in nonverbal cognitive ability, the group of females with ASD was less proficient than the gender-matched normally developing peer group on the same language tasks. 9.6.2

Social Skills in Females with ASD

An examination of pragmatic communication skills (RSPCS) (Burns et al., 1985) was used in the present study as an indicator of pragmatic social skills. This was undertaken to assess the proposal that females with ASD may have less-obvious impaired social skills (Frith, 2003; Kopp & Gillberg, 1992). Once again, it is difficult to evaluate the proposal of less-obvious impairment in females with ASD without knowing the standard against which such skills are compared. Whether compared to normally developing females or to males with ASD, the findings suggest that pragmatic deficits are obvious in females with the diagnosis. Relative to normally developing females, the females with ASD in the study presented with impairments in social interactions that were evident to their communicative partners. Their pragmatic skills were rated significantly lower by Significant Others than were the control females. Moreover, when compared to males with ASD, there was no significant difference in the Significant Others ratings, suggesting that their pragmatic deficits are neither more obvious nor less obvious than the deficits observed in males with ASD.

204 Although the findings of an earlier study in this thesis noted an accurate selfperception of pragmatic competence in the ASD group of participants (see section 6.6.1), the findings of the current study suggest self-perception of pragmatic skills, but not actual pragmatic skills, may differ between the sexes in ASD. Females with ASD rated themselves no differently to control females when questioned about their pragmatic competence. In contrast, males with ASD rated themselves significantly lower than control males on responses to pragmatic skills. Kopp and Gillberg (1992) have suggested that females with ASD may use imitative skills in social interactions. While the use of social imitation was not the focus of the current study, a small number of female (but not male) ASD participants reported imitating others in social interactions. If females with ASD use social imitation, the use of imitative skills by the female group in the current study fails to elevate their pragmatic social skills to a standard similar to that of female controls. However, if females with ASD knowingly use imitative skills in social interactions, this may explain why they think their pragmatic social skills are no different to those upon whom they model themselves. Frith (2003) has suggested that young females with ASD may have relatively intact language skills, use compensatory learning styles, and be compliance in school settings. The findings of the present study suggest that the female adults with ASD investigated in the current study do not have intact language skills. A concern, however, is that language difficulties in females with ASD could be overlooked if indeed they are more compliant and resort to compensatory learning styles (Kopp & Gillberg, 1992). 9.7

Specific Limitations of the Study

The psychometric properties of the RHLB (Bryan, 1989) and the RSPCS (Burns et al., 1985) are not known. The lack of measures of reliability and validity regarding these assessments needs to be considered when interpreting the findings of the study. Nevertheless, the subjective RSPCS gave a good indication of pragmatic skills in the current study. Future research, however, should focus on objective video analyses of

205 pragmatic skills associated with social interactions. The proposal that social imitation may be used by females with ASD should be further assessed. Additionally, although the results of the current study suggest there are no gender differences in language and communication skills in adults with ASD, it is imperative that the study be replicated with a paediatric population. 9.8

Conclusion

The language and pragmatic social skills examined in the present study failed to differentiate between the females and the males with diagnoses of ASD. Self-perception of pragmatic social skills, however, differentiated the ASD females from the ASD males, with the ASD females rating their skills on a similar level of proficiency as their typical female peers. The findings suggest, nevertheless, that females with ASD, as well as males, may benefit from communicative support.

206

10 Conclusions 10.1

Summary and Conclusions

The review of the literature presented in chapter 1 of this thesis identified three limitations of research to date which may be hindering the clinical appreciation of the language and communication difficulties experienced by individuals with a diagnosis of ASD and average intelligence. The series of studies included in the thesis, therefore, served to extend the clinical knowledge of the language and communication skills associated with the disorder. The studies reported in chapters 2 - 5 aimed to address the limitation of restricted assessments used in previous studies to define the language and communication skills of children with ASD. Chapters 6 - 8 presented studies aimed at addressing the limited clinical knowledge of the language and communication skills associated with ASD in adulthood. The study presented in chapter 9 aimed to increase the clinical understanding of the language and social skills in females with a diagnosis of ASD. 10.1.1 Child Studies Semantic and syntactic skills have been the focus of much research into AS and HFA. Theoretically, the focus on these early developing language skills aligns with DSM-IV’s (APA, 1994) language criterion and, therefore, has diagnostic relevance. Knowledge of competence in these skills, however, has limited clinical relevance, as it is the competence in skills that emerge later in the sequence of language development that may reduce psychosocial, behavioural, and academic outcomes. An extensive language/communication battery was administered to 20 children with a diagnosis of ASD. The children with ASD involved in the series of studies presented with a broad range of language deficits that extended beyond mere semantic and syntactic difficulties. Competence in overall language skills, receptive and expressive language, language content, and language memory, metalinguistic skills such as interpreting ambiguity,

207 understanding inferential language, and linguistic flexibility, prosodic functions such as the use of emphatic stress to convey contrastive meaning, and problem solving in social situations was reduced relative to their typically developing peers. The three theoretical approaches discussed in chapter 1 were applied in all but one study in the series of investigations involving the children with ASD. Findings from the validation studies suggest that AS and HFA have similar language outcomes on all but two language measures. Children reclassified as AS were more proficient than those reclassified as HFA on tasks requiring resolution of ambiguity and interpretation of metaphors presented pictorially. The validation approach to defining language and communication skills in AS and HFA provided findings that may have limited clinical applicability, however, as the approach relies upon the retrospective recall of developmental language milestones, which may be inaccurate or, as was the case for two of the children involved in the study, unavailable. The second approach, where AS and HFA were conceptualised as similar autistic presentations, revealed clinically useful information on the performance of the ASD group relative to controls. Heterogeneity of skills across the studies was noted, but this conceptualisation provided no analysis of the range of skill within the combined ASD group. The third approach, where AS and HFA were conceptualised, not as separate, or conversely, similar diagnostic categories, but as disorders on an autism spectrum, provided information that could be applied to formulate individual intervention plans for children with ASD. For example, the ability to interpret figurative language was not identified as significantly impaired in the non-differentiated ASD group of children, yet children in two of the three ASD subgroups experienced considerable difficulty with everyday figures of speech. Viewing ASD as a dimensional disorder on an autism spectrum may identify specific strengths and weaknesses in children on the spectrum, such as the case with figurative language, which can be incorporated into individualised intervention planning. Performance across all language measures was consistently high or consistently low for some of the children with ASD. Performance by ASD child participants 2, 3, 4, and 13, for instance, consistently placed these children in the most skilled ASD

208 subgroups across all language measures investigated, while performance across all language measures consistently placed ASD child participants 5 and 12 in the least competent subgroups. There was, however, evidence of uneven development of language skill across the measures for a number of the ASD children. Performance on the CELF-4 (Semel et al., 2003) by ASD child participant 9, as an example, placed this child in the second least competent subgroup on basic language skills, yet he was placed in the most competent subgroup based on his metalinguistic skills. The ASD children consistently placed in the most competent subgroups throughout the series of studies (ASD participants 2, 3, 4, and 13) presented with differing developmental language histories. Language onset for participants 2, 3, and 4 was prior to age 2 years, while the onset of language for participant 13 was delayed (after age 2 years) and speech and language intervention had been instigated some years prior his involvement in the present research. In contrast, despite presenting with a developmental history of the onset of language being prior to age 2 years, ASD participant 5 required language intervention when younger, and was consistently placed in the least competent ASD subgroups. The findings suggest that irrespective of developmental language history, comprehensive language assessments of all children with a diagnosis of pervasive developmental disorder, be it AS, HFA, or ASD, should be undertaken on a regular basis throughout childhood and adolescence to reduce the risk of inappropriate and/or ineffectual intervention and support based on incorrect clinical assumptions. 10.1.2 Adult Studies ASD is characterised by impairments in language and communication, and a history of developmental language impairment may result in negative psychosocial outcomes in adulthood. Nevertheless, there is a dearth of information on the communicative skills in adults with ASD. The findings from the series of studies investigating the language and communication skills of 17 adults with a diagnosis of ASD presented in chapters 6 - 8 of this thesis suggest that the communicative difficulties experienced by the adults may extend beyond the semantic and syntactic deficits described in previous research studies.

209 The adults with ASD presented with a range of compromised language and communication skills. Proficiency in overall linguistic ability, auditory verbal comprehension, naming skills, understanding non-contextual language such as resolving ambiguity and understanding inferential and figurative language, linguistic flexibility, social problem solving, and producing variations in emphatic stress to convey meaning in speech was reduced compared to adults with a typical developmental history. Social pragmatic skills were rated significantly lower for the ASD group than the control group for both the self-rating and the rating by a Significant Other. Eighty-two percent (14/17) of the adults with ASD were diagnosed as adults, and only 53 % (9/17) were able to provide a developmental language history. The resultant data set, which was to be used to reclassify the ASD adults into AS or HFA, was deemed too unreliable for a validation approach to defining the language and communication skills in adults with a diagnosis of ASD. Hence, as discussed throughout this thesis, the validation approach is likely to have limited clinical application, particularly in cases where diagnosis is not obtained until adulthood. The theoretical approach of combining AS, HFA, and/or ASD into one experimental group was applied in all three investigations of the adults with ASD, while the theoretical approach of the spectrum studies was applied in two of the three studies only. The findings suggest that language and communication skills within ASD are not homogeneous. Some adults in the study presented with average skills, while others experienced considerable difficulty on the tasks presented. An examination of individual differences within ASD, as in the spectrum approach, may provide clinically useful information on an individual’s communicative needs. For example, as a group, the adults with ASD presented with compromised overall linguistic skills, suggesting that linguistic support may be necessary to optimise psychosocial wellbeing. Linguistic support, however, may not necessarily be required for all adults with ASD, as two adults with ASD involved in the study presented with above-average linguistic skills. Such information can be utilised for the development of intervention services specific to the individual.

210 In addition to heterogeneity within the group of adults with ASD, there was evidence of heterogeneity of communication skills within the individual. The performance of ASD adult participants 1 and 5 placed these adults consistently in the most skilled subgroups across all language measures, and participants 4, 6, and 17 were consistent in being placed in the least competent subgroups across all measures. However, despite being placed in the most competent subgroup on linguistic/pragmatic measures, ASD adult participant 16 was placed in the least competent subgroup for metalinguistic competence. A clinical appreciation of the diversity of language and communication skills within the individual presentation, as well as within the diagnoses of AS, HFA, or ASD can only be gained by undertaking comprehensive language and communication assessments of all who present with ASD. 10.1.3 Gender Study The findings from chapters 6 - 8 indicated that, as a group, the 17 ASD adult participants presented with a range of impaired language and communication skills relative to the control group. The methodology used in chapter 9 was, effectively, within-group analyses undertaken to extend the findings from the earlier chapters focusing on the adult presentation of ASD. Due to the unreliability of developmental histories provided, coupled with the small number of participants involved, only one of the three theoretical approaches, that of viewing AS and HFA as similar autistic presentations, was applied in the study examining gender differences on language and communication tasks. Nonetheless, this level of analysis provided clinically relevant, albeit preliminary, research findings. Within-group analyses identified that the females with ASD were less proficient than the control females on overall linguistic skill, comprehension of inferred meaning, social problem solving, and production of variations in stress used to convey contrastive meaning. Further, the Significant Others’ rating of the pragmatic communication skills of the ASD females, such as nonverbal communication, conversational skills, use of linguistic context, and discourse organisation, was significantly lower than the Significant Others’ rating for the control females, suggesting impaired skill in the ASD females.

211 Moreover, the findings from chapter 9 suggest that the nine females with ASD were neither more, nor less, competent than the eight males with ASD on an extensive assessment battery comprised of 12 language tasks. Likewise, pragmatic communication deficits in the females with ASD were as obvious to Significant Others as the pragmatic deficits were in the males with the disorder. On the other hand, however, the selfperception of pragmatic communication skills differentiated the ASD females from the ASD males, with the females rating their pragmatic skills to a similar standard as the control females. Consequently, the findings suggest that females with ASD may present with impaired language and social skills, and that the impairment in females may be quantitatively no different to the impairment present in males with the disorder. The overall findings from this thesis suggest that the theoretical framework underpinning validation studies may have limited clinical (and research) utility due to the unreliability of the core premise upon which the framework is based. Conceptualisation of AS and HFA as similar autistic disorders may likewise provide only limited clinically applicable findings, as potential differences in presentation, and hence intervention needs within AS and HFA, may be overlooked. Conceptualisation of AS and HFA as dimensions on an autism spectrum, rather than distinct diagnostic categories, may have more useful application, both in research and clinical practice. 10.2

Limitations of the Present Research

While the series of studies presented in this thesis have provided valuable clinical insight into the language and communication skills associated with ASD, there are limitations of the research. Specific limitations of the individual studies have been discussed in their relevant sections, yet there are general limitations pertaining to the current research which need to be acknowledged. The criteria for diagnosing the ASD participants were not known. A number of internationally recognised classification systems are available (e.g., APA, 1994; WHO, 1992), but findings by Tager-Flusberg (2003), Klin et al. (2005), and Woodbury-Smith et al. (2005) undermine the confidence placed on the adherence to criteria for diagnostic

212 purposes. If the diagnostic criteria used to differentially diagnose the participants involved in the study were known, a lack of certainty in a strict adherence to diagnostic criteria would undermine the confidence placed on such knowledge. There are a number of instruments, such as the ADI-R (Lord et al., 1994), the ADOS-G (Lord et al., 2000), and the Autism Diagnostic Observation Schedule-WPS (Lord, Rutter, DiLavore, & Risi, 1999) that may be used in research and clinical settings to validate diagnoses on the autism spectrum. The use of such validation instruments requires “training in both administration and scoring by a person experienced in use of the instrument who has established reliability with other experienced individuals” (Lord et al., 1994, p. 682). In the current research, instruments to validate the ASD participants’ diagnoses were not utilised, as there was limited availability in the geographical region in which the study was undertaken of suitably trained personnel to administer the instruments. It is, therefore, possible that the participants in the study diagnostically classified as AS, AD, or ASD may, in fact, be incorrectly diagnosed, which would affect the results of the study. The timing of language onset and development is critical to diagnosing AD and AS. DSM-IV’s (APA, 1994) language criterion stipulates both the timing of the onset of first words and the timing of subsequent use of simple communicative phrases. The series of studies presented in this thesis utilised the timing of the onset of single words only due to parental difficulties with specifying the timing of the emergence of communicative phrases. It is possible that the lack of progress in language development from single words to communicative phrases (Chawarska & Volkmar, 2005) is pertinent in determining the diagnostic distinction between AS and HFA. The focus on onset rather than progress in skill may have influenced the findings of the studies in which a validation approach to defining the language and communication skills associated with AS and HFA was utilised. Methodological problems, such as the reliance upon the retrospective recall of developmental language milestones for the differentiation of AS from HFA (APA, 1994), inconsistency in the interpretation of milestones (Tager-Flusberg, 2003; Woodbury-Smith

213 et al., 2005), the interchangeability of the terms AS and HFA in research and clinical practice, and the inconsistent application of the language criterion in diagnostic settings (Eisenmajer et al., 1996; Klin et al., 2005) are inherent in research examining the differentiation of AS from HFA. The methodology applied in the current research adhered to DSM-IV’s language criterion relating to the onset of words to reclassify the ASD participants into AS and HFA. This methodological approach consistently interpreted and applied the reported language milestones, thus reducing a number of the inherent methodological problems noted in previous research. The reclassification of the ASD participants into AS and HFA in the present series of studies, however, still relied on parental recall of the timing of language onset in infancy, which may have been inaccurate, or not recalled. It could be argued that obtaining information regarding early language development directly from diagnosticians’ reports may be preferable to asking parents to recall such details. This would not provide more accurate data for the present study, as the mean age at diagnosis for the children was 7.9 years (age range: 3 – 12 years) and 82 % (14/17) of the adults were diagnosed in adulthood. Diagnostic reports for the majority of participants in the present study would, therefore, be based on retrospective reporting. The non-significant results in the determination of language differences between the children with differing language onset histories may reflect unreliable reporting of developmental histories. Moreover, the reliance upon retrospective recall of language milestones precluded an examination in the adult studies of differences based on language history, as many of the adults involved in the study could not provide information regarding their onset of language. While the focus of diagnostic criteria remains on retrospective language history, this issue will continue to produce unreliable research findings. This may be a particular limitation in studies involving adults with ASD, as many reach adulthood before a diagnosis is given (Tantam, 2003). As many statistical procedures rely on scores being variable and evenly distributed, ceiling effects of participants’ performance can influence findings. The effect of scores being limited by a ceiling score needs to be acknowledged in the present series

214 of studies. This is particularly relevant to the performance by the adult control participants. The box plots of their performance on the WAB (Kertesz, 1982) and a number of the RHLB (Bryan, 1989) subtests, for instance, suggest the ceiling scores may have been reached. The effect of ceiling scores on the statistical analyses needs, therefore, to be considered in interpreting the findings. Given the sample size of the study, generalising the significant results of the studies to the wider ASD population is not recommended. Replication studies, with a larger cohort and more rigorously defined samples, are required to substantiate the current findings. Furthermore, caution is required when interpreting the non-significant findings. This is particularly relevant in regards to the investigations into the differentiation of AS from HFA, the examination of the influence of nonverbal cognitive skills in subgroup formation, and the examination of the female presentation. The small number of participants involved in these aspects of the thesis may have reduced the power of the statistical analyses to detect significant differences. The use of cluster analyses considerably reduced the sample sizes, and hence, serves to highlight the heterogeneity within the ASD presentation only. Participants with a range of diagnoses on the autism spectrum, such as AS and HFA, were investigated in the present research. There is ongoing debate as to whether AS and HFA are similar autistic presentations (e.g., Frith, 2003; Klin et al., 2005; Klin & Volkmar, 2003; Leekam et al., 2000; Macintosh & Dissanayake, 2004), and it is currently not clear if the diagnostic distinction results in significantly different long-term outcomes for language (Howlin, 2003). The findings from the child studies presented in this thesis suggest that two language measures only differentiated AS from HFA. In both instances, the AS group outperformed the HFA group. As discussed above, however, the nonsignificant findings may be unreliable. The large age range (18 - 67 years for the ASD group) in the series of adult studies may have influenced the findings, as linguistic (Maxim, 1999), pragmatic (Zanini, Bryan, De Luca, & Bava, 2005), and executive function skills (Phillips, 1999) may decline with age. The high proportion of females (Fe: M = 9: 8) in the adult studies is not

215 representative of the sex ratio estimates for ASD and may have also influenced the findings of the adult studies (see McLennan, Lord, & Schopler, 1993). Due to time and financial constraints, inter-rater reliability of scoring decisions was not undertaken in the present study. A number of the subtests administered were scored objectively. Some of the subtests, however, required subjective decisions regarding the correctness of the response. Scoring guidelines from the administration manuals were adhered to, but the lack of inter-rater reliability is a limitation of the overall study. Additionally, sufficient psychometric information was not available for all assessments included in the test battery. The lack of data such as the reliability and validity of some of the measures needs to be considered when interpreting the findings of the series of studies. Additionally, nonverbal intelligence, as assessed by the TONI-2 (Brown et al., 1990), rather than full scale IQ was undertaken due to financial constraints. It is acknowledged that test scores from the TONI-2 are not overall markers of intelligence. Due to time and financial constraints, however, no IQ testing was undertaken for the study. Further, participants in the present study had no previously documented intellectual impairment. The findings, therefore, cannot be generalised to the wider ASD population where there is often an associated degree of Mental Retardation (APA, 2000). The present research focused on a cross-section of the ASD population. While it appears that language competence in ASD may be sub-optimal in childhood and in adulthood, the maturational changes in language skill from childhood to adulthood have not been examined in this thesis. Longitudinal studies would be preferable to crosssectional ones, particularly if wanting to document the developmental trajectories of children with ASD. 10.3

Future Research Directions

Nonverbal cognitive skill differentiated the ASD child group from the child control group, but not the ASD adult group from the adult control group. Nevertheless,

216 performance by the group of children with ASD and the group of adults with ASD differed significantly on a number of language tasks relative to their respective control group. Further, nonverbal cognitive skill was an inconsistent factor in the formation of subgroups based on current language performance in the child and adult studies presented in this thesis. The findings of the series of studies presented in this thesis suggest an investigation of the relationship between cognition, as indicated by full scale IQ, and language is warranted. Future research should examine whether cognitive performance, rather than the timing of language onset, is a marker that differentiates individuals with ASD. Further research is required to determine the relationship between language skill and vocational outcomes in adults with ASD. Language and communication support is often included in educational support packages for children with the disorder. Research is needed to determine if language and communication intervention and support should be included in ASD-specific vocational programmes offered to adolescents and adults on the spectrum. A number of research groups have considered structural and/or anomalous cerebral functioning in autistic presentations (e.g., Boylan, Blue, & Hohmann, 2007; Gunter et al., 2002; Ozonoff & Miller, 1996; Sabbagh, 1999). The findings from the adult studies included in the present thesis were generally based on performance on tests designed for the neurologically impaired population. Further investigation, such as the comparison of the language profiles associated with ASD and focal brain injury, may provide clinically useful data in terms of intervention strategies if the language difficulties experienced in ASD are clinically similar to those experienced following focal neurological trauma. Future research should investigate whether the level of self-perception of pragmatic communication skills noted in the adult ASD group of participants investigated in the present study facilitates positive pragmatic intervention outcomes. Studies are also needed to determine if or when children with ASD develop an accurate self-perception regarding pragmatic competence.

217 While comprehensive, the assessment battery utilised in the studies detailed in this thesis was not exhaustive. Nonverbal communication skills such as eye contact and facial expression and skills in conversation were not objectively assessed. Video analysis of these skills is required to complement the current research findings. The series of studies comprising this thesis have added to the clinical knowledgebase regarding the language and communication skills of children, adults, and females with a diagnosis of ASD. Based on the findings of the studies reported, it is concluded that the most reliable and clinically relevant method of conceptualising pervasive developmental disorders such as AS, HFA, and/or ASD is to view them as presentations on an autism spectrum. Additionally, reliance on accurate assessment of current language performance, rather than reliance on possibly inaccurate recall of past language performance should form the basis for intervention strategies and services. Hence, comprehensive language and communication assessments should be undertaken with all individuals with an autistic diagnosis, irrespective of the given diagnosis and developmental language history. As highlighted above, the findings of the studies have not incontrovertibly defined these skills in the disorders. Further research is needed to fulfil Gillberg’s (1998) first management guideline for AS and HFA, which, in part, alludes to “optimal understanding and [the provision of optimal] service” (p. 208) for individuals with a diagnosis on the autism spectrum.

218

References Adams, C., Green, J., Gilchrist, A., & Cox, A. (2002). Conversational behaviour of children with Asperger syndrome and conduct disorder. Journal of Child Psychology and Psychiatry, 43(5), 679-690. Aldenderfer, M. S., & Blashfield, R. K. (1984). Cluster analysis. Beverly Hills, CA: Sage Publications. Alpha Autism Inc. (2007). Adult training and support services. Retrieved February 1, 2007, from http://www.autismvictoria.org.au/about/organisational_members.php# American Psychiatric Association [APA]. (1980). Diagnostic and statistical manual of mental disorders (3rd ed.). Washington, DC: Author. American Psychiatric Association [APA]. (1987). Diagnostic and statistical manual of mental disorders (3rd ed.-Rev.). Washington, DC: Author. American Psychiatric Association [APA]. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author. American Psychiatric Association [APA]. (2000). Diagnostic and statistical manual of mental disorders (4th ed. Text Revision). Washington, DC: Author American Speech-Language-Hearing Association. (2000). Guidelines for the roles and responsibilities of the school-based speech-language pathologist. Rockville, MD: Author. American Speech-Language-Hearing Association. (2006). Guidelines for speechlanguage pathologists in diagnosis, assessment, and treatment of autism spectrum disorders across the lifespan. Available from http://www.asha.org/members/deskref-journal/deskref/default. Retrieved June 3, 2006 from source.

219 Asperger, H. (1991). 'Autistic psychopathy' in childhood (U. Frith, Trans.). In U. Frith (Ed.), Autism and Asperger syndrome (pp. 37-92). Cambridge: Cambridge University Press. Attwood, T. (1998). Asperger's syndrome: A guide for parents and professionals. London: Jessica Kingsley. Bachevalier, J. (1994). Medial temporal lobe structures and autism: A review of clinical and experimental findings. Neuropsychologia, 32(6), 627-648. Baldo, J. V., Dronkers, N. F., Wilkins, D., Ludy, C., Raskin, P., & Kim, J. (2005). Is problem solving dependent on language? Brain and Language, 92, 240-250. Baron-Cohen, S. (1988). Social and pragmatic deficits in autism: Cognitive or affective? Journal of Autism and Developmental Disorders, 18(3), 379-402. Baron-Cohen, S. (2002). The extreme male brain theory of autism. Trends in Cognitive Sciences, 6(6), 248-254. Baron-Cohen, S., Jolliffe, T., Mortimore, C., & Robertson, M. (1997). Another advanced test of theory of mind: Evidence from very high functioning adults with autism or Asperger syndrome. Journal of Child Psychology and Psychiatry, 38(7), 813-822. Baxter, M. J. (1994). Exploratory multivariate analysis in archeology. Edinburgh: Edinburgh University Press. Bishop, D. (1997). Uncommon understanding: Development and disorders of language comprehension in children. Hove: Psychology Press. Bishop, D. (1998). Development of the Children's Communication Checklist (CCC): A method for assessing qualitative aspects of communicative impairment in children. Journal of Child Psychology and Psychiatry, 39, 879-891. Bishop, D. (2003). The Children's Communication Checklist (2nd ed.). London: The Psychological Corporation.

220 Bishop, D. (2006, February). Exploring the boundaries of autistic disorder. Paper presented at the Communication Disorders in Childhood Conference, Children's Hospital Education Research Institute, Westmead, Sydney, Australia. Bloom, L., & Lahey, M. (1978). Language development and language disorders. New York: John Wiley & Sons. Booth, R., Charlton, R., Hughes, C., & Happe, F. (2003). Disentangling weak coherence and executive dysfunction: Planning drawing in autism and attention deficit/hyperactivity disorder. Philosophical Transactions of the Royal Society of London B, 358, 387-392. Bottini, G., Corcoran, R., Sterzi, R., Paulesu, E., Schenone, P., Scarpa, P., et al. (1994). The role of the right hemisphere in the interpretation of figurative aspects of language: A positron emission tomography activation study. Brain, 117, 12411253. Bowers, L., Barrett, M., Huisingh, R., Orman, J., & LoGiudice, C. (1991). Test of Problem Solving-Adolescent. East Moline, IL: LinguiSystems Inc. Boylan, C. B., Blue, M. E., & Hohmann, C. F. (2007). Modeling early cortical serotonergic deficits in autism. Behavioural Brain Research, 176(1), 94-108. Bradshaw, J. L. (2001). Developmental disorders of the frontostriatal system. East Sussex: Psychology Press. Brown, L., Sherbenou, R. J., & Johnsen, S. K. (1990). Test of Nonverbal Intelligence (2nd ed.). Austin, TX: Pro-Ed. Bryan, K. L. (1989). The Right Hemisphere Language Battery. Leicester, England: Far Communications.

221 Burns, M. S., Halper, A. S., & Mogil, S. I. (1985). Rating Scale of Pragmatic Communication Skills. In M. S. Burns, A. S. Halper, & S. I. Mogil (Eds.), Clinical management of right hemisphere dysfunction (pp. 50-52). Rockville, MD: Aspen Systems Corporation. Channon, S. (2004). Frontal lobe dysfunction and everyday problem-solving: Social and non-social contributions. Acta Psychologica, 115, 235-254. Channon, S., Charman, T., Heap, J., Crawford, S., & Rios, P. (2001). Real-life-type problem-solving in Asperger's syndrome. Journal of Autism and Developmental Disorders, 31(5), 461-469. Channon, S., & Crawford, S. (1999). Problem-solving in real-life-type situations: The effects of anterior and posterior lesions on performance. Neuropsychologia, 37, 757-770. Charman, T. (2004). Ask the editor. Journal of Autism and Developmental Disorders, 34(4), 467. Chawarska, K., & Volkmar, F. R. (2005). Autism in infancy and early childhood. In F. R. Volkmar, R. Paul, A. Klin, & D. Cohen (Eds.), Handbook of autism and pervasive developmental disorders (3 rd ed., Vol. 1) (pp. 223-246). Hoboken, NJ: John Wiley & Sons. Clegg, J., Hollis, C., Mawhood, L., & Rutter, M. (2005). Developmental language disorders - a follow-up in later adult life. Cognitive, language and psychosocial outcomes. Journal of Child Psychology and Psychiatry, 46(2), 128-149. Croen, L. A., Grether, J. K., Hoogstrate, J., & Selvin, S. (2002). The changing prevalence of autism in California. Journal of Autism and Developmental Disorders, 32(3), 207-215. Cronin, V. S. (2002). The syntagmatic-paradigmatic shift and reading development. Journal of Child Language, 29(1), 189-204.

222 Dennis, M., Lazenby, A. L., & Lockyer, L. (2001). Inferential language in high-function children with autism. Journal of Autism and Developmental Disorders, 31(1), 4754. Di Simoni, F. (1977). The Token Test for Children. Boston, MA: Teaching Resources. Doherty, M. J. (2000). Children's understanding of homonymy: Metalinguistic awareness and false belief. Journal of Child Language, 27(2), 367-392. Dunn, L., & Dunn, G. (1981). Peabody Picture Vocabulary Test (Rev. ed.). Circle Pines, MN: American Guidance Service. Dunn, L., Dunn, L., Whetton, C., & Burley, J. (1997). British Picture Vocabulary Scale (Rev. ed.). Slough, Bucks: NFER-Nelson Publishing Co. Dunn, L., Dunn, L., Whetton, C., & Pintillie, D. (1982). British Picture Vocabulary Scale. Slough, Bucks: NFER-Nelson Publishing Co. Eales, M. J. (1993). Pragmatic impairments in adults with childhood diagnoses of autism or developmental receptive language disorder. Journal of Autism and Developmental Disorders, 23(4), 593-617. Eisenmajer, R., Prior, M., Leekam, S., Wing, L., Gould, J., Welham, M., et al. (1996). Comparison of clinical symptoms in autism and Asperger's disorder. Journal of the American Academy of Child and Adolescent Psychiatry, 35(11), 1523-1531. Ellis, H. D., & Gunter, H. L. (1999). Asperger syndrome: A simple matter of white matter? Trends in Cognitive Neuroscience, 3(5), 192-200. Emerich, D. M., Creaghead, N. A., Grether, S. M., Murray, D., & Grasha, C. (2003). The comprehension of humorous materials by adolescents with high-functioning autism and Asperger's syndrome. Journal of Autism and Developmental Disorders, 33(3), 253-257.

223 Ervin, S. M. (1961). Changes with age in the verbal determinants of wordassociation. American Journal of Psychology, 74, 361-372. Everitt, B. S., Landau, S., & Leese, M. (2001). Cluster analysis (4th ed.). London: Arnold. Fine, J., Bartolucci, G., Ginsberg, G., & Szatmari, P. (1991). The use of intonation to communicate in pervasive developmental disorders. Journal of Child Psychology and Psychiatry, 32(5), 771-782. Fine, J., Bartolucci, G., Szatmari, P., & Ginsberg, G. (1994). Cohesive discourse in pervasive developmental disorders. Journal of Autism and Developmental Disorders, 24(3), 315-329. Fletcher-Watson, S., Leekam, S. R., Turner, M. A., & Moxon, L. (2006). Do people with autistic spectrum disorder show normal selection for attention? Evidence from change blindness. British Journal of Psychology, 97, 537-554. Frith, U. (1991). Asperger and his syndrome. In U. Frith (Ed.), Autism and Asperger syndrome (pp. 1-36). Cambridge: Cambridge University Press. Frith, U. (2003). Autism: Explaining the enigma (2nd ed.). Oxford: Blackwell Publishing. Gardner, M. C. (1982). Expressive One Word Picture Vocabulary Test (Upper Extension). Los Angeles: Western Psychological Services. Ghaziuddin, M., & Gerstein, L. (1996). Pedantic speaking style differentiates Asperger syndrome from high-functioning autism. Journal of Autism and Developmental Disorders, 26(6), 585-595. Ghaziuddin, M., Thomas, P., Napier, E., Kearney, G., Tsai, L., Welch, K., et al. (2000). Brief report: Brief syntactic analysis in Asperger syndrome: A preliminary study. Journal of Autism and Developmental Disorders, 30(1), 67-70.

224 Gilchrist, A., Green, J., Cox, A., Burton, D., Rutter, M., & Le Couteur, A. (2001). Development and current functioning in adolescents with Asperger syndrome: A comparative study. Journal of Child Psychology and Psychiatry, 42(2), 227-240. Gillberg, C. (1989). Asperger syndrome in 23 Swedish children. Developmental Medicine and Child Neurology, 31, 520-531. Gillberg, C. (1998). Asperger syndrome and high-functioning autism. British Journal of Psychiatry, 172, 200-209. Gillberg, C., & Rastam, M. (1992). Do some cases of anorexia nervosa reflect underlying autistic-like conditions? Behavioural Neurology, 5, 27-32. Gillberg, I. C., & Gillberg, C. (1989). Asperger syndrome-some epidemiological considerations: A research note. Journal of Child Psychology and Psychiatry, 30(4), 631-638. Goldberg, M. C., Mostofsky, S. H., Cutting, L. E., Mahone, E. M., Astor, B. C., Denckla, M. B., et al. (2005). Subtle executive impairment in children with autism and children with ADHD. Journal of Autism and Developmental Disorders, 35(3), 279-293. Griswold, D. E., Barnhill, G. P., Myles, B. S., Hagiwara, T., & Simpson, R. L. (2002). Asperger syndrome and academic achievement. Focus on Autism and Other Developmental Disabilities, 17(2), 94-102. Gunter, H. L., Ghaziuddin, M., & Ellis, H. D. (2002). Asperger syndrome: Tests of right hemisphere functioning and interhemispheric communication. Journal of Autism and Developmental Disorders, 32(4), 263-281. Hammill, D. D., Brown, V. L., Larsen, S. C., & Wiederholt, J. L. (1980). Test of Adolescent Language: A multidimensional approach to assessment. Austen, TX: Pro-Ed.

225 Hammill, D. D., Brown, V. L., Larsen, S. C., & Wiederholt, J. L. (1994). Test of Adolescent Language (3rd ed.). Austin, TX: Pro-Ed. Hill, E. L. (2004). Evaluating the theory of executive dysfunction in autism. Developmental Review, 24, 189-233. Hill, E. L., & Frith, U. (2003). Understanding autism: Insights from mind and brain. Philosophical Transactions of the Royal Society of London B, 358, 281-289. Howlin, P. (1997). Autism: Preparing for adulthood (2nd ed.). London: Routledge. Howlin, P. (2003). Outcome in high-functioning adults with autism with and without early language delays: Implications for the differentiation between autism and Asperger syndrome. Journal of Autism and Developmental Disorders, 33(1), 313. Howlin, P., Alcock, J., & Burkin, C. (2005). An 8 year follow-up of a specialist supported employment service for high-ability adults with autism or Asperger syndrome. Autism, 9(5), 533-549. Howlin, P., & Asgharian, A. (1999). The diagnosis of autism and Asperger syndrome: Findings from a survey of 770 families. Developmental Medicine and Child Neurology, 41, 834-839. Jolliffe, T., & Baron-Cohen, S. (2000). Linguistic processing in high-functioning adults with autism or Asperger's syndrome. Is global coherence impaired? Psychological Medicine, 30, 1169-1187. Joseph, R. M., McGrath, L., & Tager-Flusberg, H. (2005). Executive dysfunction and its relation to language ability in verbal school-age children with autism. Developmental Neuropsychology, 27(3), 361-378. Kadesjo, B., Gillberg, C., & Hagberg, B. (1999). Brief report: Autism and Asperger syndrome in seven-year-old children: A total population study. Journal of Autism and Developmental Disorders, 29(4), 327-331.

226 Kaland, N., Moller-Nielsen, A., Callesen, K., Mortensen, E. L., Gottlieb, D., & Smith, L. (2002). A new 'advanced' test of theory of mind: Evidence from children and adolescents with Asperger syndrome. Journal of Child Psychology and Psychiatry, 43(4), 517-528. Kanner, L. (1942-3). Autistic disturbances of affective contact. The Nervous Child, 2(1), 217-250. Kaplan, E. F., Goodglass, H., & Weintraub, S. (1978). Boston Naming Test (exper. ed.). E. Kaplan & H. Goodglass. Kertesz, A. (1982). Western Aphasia Battery. New York: Grune and Stratton. Klin, A. (2000). Attributing social meaning to ambiguous visual stimuli in higherfunctioning autism and Asperger syndrome: The Social Attribution Task. Journal of Child Psychology and Psychiatry, 41(7), 831-846. Klin, A., Pauls, D., Schultz, R., & Volkmar, F. (2005). Three diagnostic approaches to Asperger syndrome: Implications for research. Journal of Autism and Developmental Disorders, 35(2), 221-234. Klin, A., & Volkmar, F. R. (2003). Asperger syndrome: Diagnosis and external validity. Child and Adolescent Psychiatric Clinics of North America, 12, 1-13. Klin, A., Volkmar, F. R., Sparrow, S. S., Cicchetti, D. V., & Rourke, B. P. (1995). Validity and neuropsychological characterization of Asperger syndrome: Convergence with nonverbal learning disabilities syndrome. Journal of Child Psychology and Psychiatry, 36(7), 1127-1140. Klonoff, H., Clark, C., Oger, J., Paty, D., & Li, D. (1991). Neuropsychological performance in patients with mild multiple sclerosis. Journal of Nervous and Mental Disease, 179, 127-131.

227 Koning, C., & Magill-Evans, J. (2001). Social and language skills in adolescent boys with Asperger syndrome. Autism, 5(1), 23-36. Kopp, S., & Gillberg, C. (1992). Girls with social deficits and learning problems: Autism, atypical Asperger syndrome or a variant of these conditions. European Child and Adolescent Psychiatry, 1(2), 89-99. Landa, R. J., & Goldberg, M. C. (2005). Language, social, and executive functions in high functioning autism: A continuum of performance. Journal of Autism and Developmental Disorders, 35(5), 557-573. Lane, D. M. (n.d.). Box Plots. Retrieved December 6, 2006, from http://davidmlane.com/hyperstat/A37797.html LeCouteur, A., Rutter, M., Lord, C., Rios, P., Robertson, S., Holdgrafer, M., et al. (1989). Autism Diagnostic Interview: A standardized investigation-based instrument. Journal of Autism and Developmental Disorders, 19, 363-387. Leekam, S., Libby, S., Wing, L., Gould, J., & Gillberg, C. (2000). Comparison of ICD-10 and Gillberg's criteria for Asperger syndrome. Autism, 4(1), 11-28. Lethlean, J. B., & Murdoch, B. E. (1997). Performance of subjects with multiple sclerosis on tests of high level language. Aphasiology, 11, 39-57. Levorato, M. C., & Cacciari, C. (2002). The creation of new figurative expressions: Psycholinguistic evidence in Italian children, adolescents and adults. Journal of Child Language, 29(1), 127-150. Liss, M., Fein, D., Allen, D., Dunn, M., Feinstein, C., Morris, R., et al. (2001). Executive functioning in high-functioning children with autism. Journal of Child Psychology and Psychiatry, 42(2), 261-270. Lloyd, P., Mann, S., & Peers, I. (1998). The growth of speaker and listener skills from five to eleven years. First Language, 18, 81-103.

228 Lord, C. (2000). Commentary: Achievements and future directions for intervention research in communication and autism spectrum disorders. Journal of Autism and Developmental Disorders, 30(5), 393-398. Lord, C., Risi, S., Lambrecht, L., Cook, E. H. J., Leventhal, B. L., DiLavore, P. C., et al. (2000). The Autistic Diagnostic Observation Schedule-Generic: A standard measure of social and communication deficits. Journal of Autism and Developmental Disorders, 30(3), 205-223. Lord, C., Rutter, M., DiLavore, P. C., & Risi, S. (1999). Autism Diagnostic Observation Schedule-WPS. Los Angeles: Western Psychological Services. Lord, C., Rutter, M., Goode, S., Heemsburgen, J., Jordan, H., Mawhood, L., et al. (1989). Autism Diagnostic Observation Schedule: A standardized observation of communicative and social behavior. Journal of Autism and Developmental Disorders, 19, 185-212. Lord, C., Rutter, M., & LeCouteur, A. (1994). Autism Diagnostic Interview-Revised: A revised version of a diagnostic interview for caregivers of individuals with possible pervasive developmental disorders. Journal of Autism and Developmental Disorders, 24, 659-685. Losh, M., & Capps, L. (2003). Narrative ability in high-functioning children with autism or Asperger's syndrome. Journal of Autism and Developmental Disorders, 33(3), 239-251. Lublin, F. D., & Reingold, S. C. (2003). Clinical features and subtypes of multiple sclerosis. In W. I. McDonald & J. H. Noseworthy (Eds.), Multiple sclerosis 2 (pp. 13-20). Philadelphia: Butterworth-Heinemann. Macintosh, K. E., & Dissanayake, C. (2004). Annotation: The similarities and differences between autistic disorder and Asperger's disorder: A review of the empirical evidence. Journal of Child Psychology and Psychiatry, 45, 421-434.

229 Mandell, D. S., Walrath, C. M., Manteuffel, B., Sgro, G., & Pinto-Martin, J. (2005). Characteristics of children with autistic spectrum disorders served in comprehensive community-based mental health settings. Journal of Autism and Developmental Disorders, 35(3), 313-321. Manjiviona, J. (2003). Assessment of specific learning difficulties. In M. Prior (Ed.), Learning and behavior problems in Asperger syndrome (pp. 55-84). New York: Guilford Press. Martin, I., & McDonald, S. (2004). An exploration of causes of non-literal language problems in individuals with Asperger syndrome. Journal of Autism and Developmental Disorders, 34(3), 311-328. Maurer, R. G., & Damasio, A. R. (1982). Childhood autism from the point of view of behavioral neurology. Journal of Autism and Developmental Disorders, 12(2), 195-205. Mawhood, L., & Howlin, P. (1999). The outcome of a supported employment scheme for high-functioning adults with autism or Asperger syndrome. Autism, 3(3), 229-254. Mawhood, L., Howlin, P., & Rutter, M. (2000). Autism and developmental receptive language disorder: A comparative follow-up in early adult life. I: Cognitive and language outcomes. Journal of Child Psychology and Psychiatry, 41(5), 547-559. Maxim, J. (1999). Ageing and language. In F. Fabbro (Ed.), Concise encyclopedia of language pathology (pp. 142-146). Amsterdam: Elsevier. Mayes, S. D., & Calhoun, S. L. (1999). Symptoms of autism in young children and correspondence with the DSM. Infants and Young Children, 12, 90-97. Mayes, S. D., & Calhoun, S. L. (2001). Non-significance of early speech delay in children with autism and normal intelligence and implications for DSM-IV Asperger's disorder. Autism, 5(1), 81-94.

230 Mayes, S. D., & Calhoun, S. L. (2003). Relationship between Asperger syndrome and high functioning autism. In M. Prior (Ed.), Learning and behavior problems in Asperger syndrome (pp. 15-34). New York: Guilford Press. Mayes, S. D., Calhoun, S. L., & Crites, D. L. (2001). Does DSM-IV Asperger's disorder exist? Journal of Abnormal Child Psychology, 29(3), 263-271. McCann, J., & Peppe, S. (2003). Prosody in autism spectrum disorders: A critical review. International Journal of Language and Communication Disorders, 38(4), 325350. McCarthy, D. (1972). McCarthy Scales of Children's Ability. New York: Psychological Corporation. McLennan, J. D., Lord, C., & Schopler, E. (1993). Sex differences in higher functioning people with autism. Journal of Autism and Developmental Disorders, 23(2), 217227. Mesibov, G. B., Shea, V., & Adams, L. W. (2001). Understanding Asperger syndrome and high functioning autism. New York: Kluwer Academic/Plenum Publishers. Miller, J. N., & Ozonoff, S. (1997). Did Asperger's cases have Asperger disorder? Journal of Child Psychology and Psychiatry, 38(2), 247-251. Miller, J. N., & Ozonoff, S. (2000). The external validity of Asperger disorder: Lack of evidence from the domain of neuropsychology. Journal of Abnormal Child Psychology, 109(2), 227-238. Minshew, N. J., Goldstein, G., Muenz, L. R., & Payton, J. B. (1992). Neuropsychological functioning in nonmentally retarded autistic individuals. Journal of Clinical and Experimental Neuropsychology, 14(5), 749-761.

231 Minshew, N. J., Siegel, D. J., Goldstein, G., & Weldy, S. (1994). Verbal problem solving in high functioning autistic individuals. Archives of Clinical Neuropsychology, 9, 31-40. Mosher, F. A., & Hornsby, J. R. (1966). On asking questions. In J. S. Bruner, R. R. Olver, T. M. Greenfield, J. R. Hornsby, H. J. Kenney, M. Maccoby, et al. (Eds.), Studies in cognitive growth (pp. 86-102). New York: John Wiley & Sons. Murdoch, B. E., & Lethlean, J. B. (2000). High-level language, naming and discourse abilities in multiple sclerosis. In B. E. Murdoch & D. G. Theodoros (Eds.), Speech and language disorders in multiple sclerosis (pp. 131-154). London: Whurr Publishers. National Autistic Society [NAS]. (2006). Prospects. Retrieved May 9, 2006, from http://www.nas.org.uk/nas/jsp/polopoly National Autistic Society [NAS]. (2007). NAS adult services-An Introduction. Retrieved February 1, 2007, from http://www.nas.org.uk/nas/jsp/polopoly.jsp?d=139 NCSU. (n.d.). Cluster analysis. Retrieved November 29, 2006, from http://www2.chass.ncsu.edu/garson/PA765/cluster.htm. Newcomer, P. L., & Hammill, D. D. (1988). Test of Language Development-2 Primary. Austin, TX: Pro-Ed. Ninio, A. (2004). Young children's difficulty with adjectives modifying nouns. Journal of Child Language, 31(2), 255-285. Njiokiktjien, C., Verschoor, A., de Sonneville, L., Huyser, C., Op het Veld, V., & Toorenaar, N. (2001). Disordered recognition of facial identity and emotions in three Asperger type autists. European Child and Adolescent Psychiatry, 10(1), 7990.

232 Noens, I. L. J., & van Berckelaer-Onnes, L. A. (2005). Captured by details: Sensemaking, language and communication in autism. Journal of Communication Disorders, 38, 123-141. Nyden, A., Billstedt, E., Hjelmquist, E., & Gillberg, C. (2001). Neurocognitive stability in Asperger syndrome, ADHD, and reading and writing disorder: A pilot study. Developmental Medicine and Child Neurology, 43, 165-171. Olver, R. R., & Hornsby, J. R. (1966). On equivalence. In J. S. Bruner, R. R. Olver, T. M. Greenfield, J. R. Hornsby, H. J. Kenney, M. Maccoby, et al. (Eds.), Studies in cognitive growth (pp. 68-85). New York: John Wiley & Sons. Ozonoff, S., & Miller, J. N. (1996). An exploration of right-hemisphere contributions to the pragmatic impairments of autism. Brain and Language, 52, 411-434. Ozonoff, S., South, M., & Miller, J. N. (2000). DSM-IV-defined Asperger syndrome: Cognitive, behavioral and early history differentiation from high-functioning autism. Autism, 4(1), 29-46. Palermo, D. S. (1971). Characteristics of word association responses obtained from children in grades one through four. Developmental Psychology, 5(1), 118-123. Paul, R., Augustyn, A., Klin, A., & Volkmar, F. (2005). Perception and production of prosody by speakers with autism spectrum disorders. Journal of Autism and Developmental Disorders, 35(2), 205-220. Pennington, B. F., & Ozonoff, S. (1996). Executive functions and developmental psychopathology. Journal of Child Psychology and Psychiatry, 37(1), 51-87. Phillips, L. H. (1999). Age and individual differences in letter fluency. Developmental Neuropsychology, 15, 249-267.

233 Pomeroy, J. C. (1998). Subtyping pervasive developmental disorder: Issues of validity and implications for child psychiatric diagnosis. In E. Schopler, G. B. Mesibov, & L. J. Kunce (Eds.), Asperger syndrome or high functioning autism? (pp. 29-60). New York: Plenum Press. Price, I. (2000). Bonferroni adjustment. Retrieved November 29, 2006, from www.une.edu.au/WebStat/unit_materials/c7_anova/oneway_bonferroni_adjust.ht m. Prior, M., & Hoffmann, W. (1990). Brief report: Neuropsychological testing of autistic children through an exploration with frontal lobe tests. Journal of Autism and Developmental Disorders, 20(4), 581-590. Prior, M., Leekam, S., Ong, B., Eisenmajer, R., Wing, L., Gould, J., et al. (1998). Are there subgroups within the autistic spectrum? A cluster analysis of a group of children with autistic spectrum disorders. Journal of Child Psychology and Psychiatry, 39(6), 893-902. Psychological Corporation. (1991). Wechsler Individual Achievement Test. New York: Author. Ramberg, C., Ehlers, S., Nyden, A., Johansson, M., & Gillberg, C. (1996). Language and pragmatic functions in school-age children on the autism spectrum. European Journal of Disorders of Communication, 31, 387-413. Rapin, I., & Allen, D. (1983). Developmental language disorders: Nosologic considerations. In U. Kirk (Ed.), Neuropsychology of language, reading, and spelling (pp. 155-184). New York: Academic Press. Reynell, J. K., & Huntley, M. (1987). Reynell Developmental Language Scales. Windsor, England: NFER-Nelson.

234 Rinehart, N. J., Bradshaw, J. L., Brereton, A. V., & Tonge, B. J. (2002). Lateralization in individuals with high-functioning autism and Asperger's disorder: A frontostriatal model. Journal of Autism and Developmental Disorders, 32(4), 321-332. Rinehart, N. J., Bradshaw, J. L., Moss, S. A., Brereton, A. V., & Tonge, B. J. (2001). A deficit in shifting attention present in high-functioning autism but not Asperger's disorder. Autism, 5(1), 67-80. Rinehart, N. J., Bradshaw, J. L., Tonge, B. J., Brereton, A. V., & Bellgrove, M. A. (2002). A neurobehavioral examination of individuals with high-functioning autism and Asperger's disorder using a fronto-striatal model of dysfunction. Behavioral and Cognitive Neuroscience Reviews, 1(2), 164-177. Rouse, R. (2004). Autism 'labels' are falsified, study finds. Retrieved April 3, 2006, from http://www.medicalobserver.com.au/displayarticle/index.asp?articleID=2686&te mplateID=105&url=http://Lock:87%2Fsearch%2Fsearch%2Easp?keywords%3D 13%2BJune%2B2003 Rumsey, J. M. (1985). Conceptual problem-solving in highly verbal, nonretarded autistic men. Journal of Autism and Developmental Disorders, 15(1), 23-36. Rumsey, J. M., & Hamburger, S. D. (1988). Neuropsychological findings in highfunctioning men with infantile autism, residual state. Journal of Clinical and Experimental Neuropsychology, 10(2), 210-221. Russell, J., & Jarrold, C. (1998). Error-correction problems in autism: Evidence for a monitoring impairment? Journal of Autism and Developmental Disorders, 28(3), 177-188. Rutherford, M. D., Baron-Cohen, S., & Wheelwright, S. (2002). Reading the mind in the voice: A study with normal adults and adults with Asperger syndrome and high functioning autism. Journal of Autism and Developmental Disorders, 32(3), 189194.

235 Rutter, M. (1978). Diagnosis and definition of childhood autism. Journal of Autism and Childhood Schizophrenia, 8(2), 139-161. Rutter, M. (2005). Autism research: Lessons from the past and prospects for the future. Journal of Autism and Developmental Disorders, 35(2), 241-257. Sabbagh, M. A. (1999). Communicative intentions and language: Evidence from righthemisphere damage and autism. Brain and Language, 70, 29-69. Schopler, E. (1985). Convergence of learning disability, higher-level autism, and Asperger's syndrome. Journal of Autism and Developmental Disorders, 15(4), 359-360. Schopler, E., Reichler, R. J., & Renner, B. R. (1988). The Childhood Autism Rating Scale (CARS). Los Angeles: Western Psychological Services. Scott, D. W. (1985). Asperger's syndrome and non-verbal communication: A pilot study. Psychological Medicine, 15, 683-687. Sekerina, I. A., Stromswold, K., & Hestvik, A. (2004). How do adults and children process referentially ambiguous pronouns? Journal of Child Language, 31(1), 123-152. Seltzer, M. M., Krauss, M. W., Shattuck, P. T., Orsmond, G., Swe, A., & Lord, C. (2003). The symptoms of autism spectrum disorders in adolescence and adulthood. Journal of Autism and Developmental Disorders, 33(6), 565-581. Semel, E., Wiig, E. H., & Secord, W. (1987). Clinical Evaluation of Language Fundamentals (Rev. ed.). San Antonio, CA: Psychological Corporation. Semel, E., Wiig, E. H., & Secord, W. A. (2003). Clinical Evaluation of Language Fundamentals (4th ed.). New York: The Psychological Corporation.

236 Semrud-Clikeman, M., & Hynd, G. W. (1990). Right hemispheric dysfunction in nonverbal learning disabilities: Social, academic, and adaptive functioning in adults and children. Psychological Bulletin, 107(2), 196-209. Shewan, C. M., & Kertesz, A. (1980). Reliability and validity characteristics of the Western Aphasia Battery (WAB). Journal of Speech and Hearing Disorders, 45, 308-324. Shields, J., Varley, R., Broks, P., & Simpson, A. (1996). Hemispheric function in developmental language disorders and high-level autism. Developmental Medicine and Child Neurology, 38, 473-486. Shriberg, L. D., Paul, R., McSweeny, J. L., Klin, A., Cohen, D. J., & Volkmar, F. R. (2001). Speech and prosody characteristics of adolescents and adults with highfunctioning autism and Asperger syndrome. Journal of Speech, Language, and Hearing Research, 44, 1097-1115. Sokal, R. R., & Michener, C. D. (1958). A statistical method for evaluating systematic relationships. University of Kansas Science Bulletin, 38(1409-1438). Solomon, M., Goodlin-Jones, B. L., & Anders, T. F. (2004). A social adjustment enhancement intervention for high functioning autism, Asperger's syndrome, and pervasive developmental disorder NOS. Journal of Autism and Developmental Disorders, 34(6), 649-668. Sperry, L. A., & Mesibov, G. B. (2005). Perceptions of social challenges of adults with autism spectrum disorder. Autism, 9(4), 362-376. Spreen, O., & Benton, A. L. (1969). Neurosensory Center Comprehensive Examination for Aphasia. Victoria, Canada: University of Victoria.

237 Starr, E., Szatmari, P., Bryson, S., & Zwaigenbaum, L. (2003). Stability and change among high-functioning children with pervasive developmental disorders: A 2year outcome study. Journal of Autism and Developmental Disorders, 33(1), 1522. StatSoft, (1984-2004). Cluster analysis. Retrieved February 16, 2006, from http://.statsoft.com/textbook/stcluan.html Stone, V., Baron-Cohen, S., & Knight, R. T. (1998). Frontal lobe contributions to theory of mind. Journal of Cognitive Neuroscience, 10(5), 640-656. Sturm, H., Fernell, E., & Gillberg, C. (2004). Autism spectrum disorders in children with normal intellectual levels: Associated impairments and subgroups. Developmental Medicine and Child Neurology, 46, 444-447. Szatmari, P., Archer, L., Fisman, S., Streiner, D. L., & Wilson, F. (1995). Asperger's syndrome and autism: Differences in behavior, cognition, and adaptive functioning. Journal of the American Academy of Child and Adolescent Psychiatry, 34(12), 1662-1671. Szatmari, P., Bartolucci, G., Finlayson, A., & Krames, L. (1986). A vote for Asperger's syndrome. Journal of Autism and Developmental Disorders, 16(4), 515-517. Szatmari, P., Bremner, R., & Nagy, J. (1989). Asperger's syndrome: A review of clinical features. Canadian Journal of Psychiatry, 34(6), 554-560. Szatmari, P., Bryson, S. E., Streiner, D. L., Wilson, F., Archer, L., & Ryerse, C. (2000). Two-year outcome of preschool children with autism or Asperger's syndrome. American Journal of Psychiatry, 157(12), 1980-1987. Szatmari, P., Tuff, L., Finlayson, A. J., & Bartolucci, G. (1990). Asperger's syndrome and autism: Neurocognitive aspects. Journal of the American Academy of Child and Adolescent Psychiatry, 29(1), 130-136.

238 Tager-Flusberg, H. (2003). Effects of language and communicative deficits on learning and behavior. In M. Prior (Ed.), Learning and behavior problems in Asperger syndrome (pp. 85-103). New York: Guilford Press. Tager-Flusberg, H. (2004). Strategies for conducting research on language in autism. Journal of Autism and Developmental Disorders, 34(1), 75-80. Tantam, D. (1991). Asperger syndrome in adulthood. In U. Frith (Ed.), Autism and Asperger syndrome (pp. 147-183). Cambridge: Cambridge University Press. Tantam, D. (2003). The challenge of adolescents and adults with Asperger syndrome. Child and Adolescent Psychiatric Clinics of North America, 12, 143-163. Tantam, D., Holmes, D., & Cordess, C. (1993). Nonverbal expression in autism of Asperger type. Journal of Autism and Developmental Disorders, 23(1), 111-133. Terman, L. M., & Merrill M. A. (1960). Stanford-Binet Intelligence Scale. Boston: Houghton Mifflin. Tompkins, C. A. (1995). Right hemisphere communication disorders: Theory and management. California: Singular Publishing Group, Inc. Verte, S., Geurts, H. M., Roeyers, H., Oosterlaan, J., & Sergeant, J. A. (2006). Executive functioning in children with autism spectrum disorder: Can we differentiate within the spectrum? Journal of Autism and Developmental Disorders, 36(3), 351-372. Verte, S., Geurts, H. M., Roeyers, H., Rosseel, Y., Oosterlaan, J., & Sergeant, J. A. (2006). Can the Children's Communication Checklist differentiate autism spectrum subtypes? Autism, 10(3), 266-287.

239 Volden, J. (2004). Conversational repair in speakers with autism spectrum disorder. International Journal of Language and Communication Disorders, 39(2), 171189. Warreyn, P., Roeyers, H., & De Groote, I. (2005). Early social communicative behaviours of preschoolers with autism spectrum disorder during interaction with their mothers. Autism, 9(4), 342-361. Wechsler, D. (1955). Wechsler Adult Intelligence Scale. New York: Psychological Corporation. Wechsler, D. (1974). Wechsler Intelligence Scale for Children-Revised. San Antonio, TX: Psychological Corporation. Wechsler, D. (1981). Wechsler Adult Intelligence Scale-Revised. San Antonio, CA: Psychological Corporation. Weinbach, R. W., & Grinnell Jr., R. M. (1998). Statistics for social workers (4th ed.). New York: Longman. Wells, B., Peppe, S., & Goulandris, N. (2004). Intonation development from five to thirteen. Journal of Child Language, 31(4), 749-778. Wetherby, A. M., & Gaines, B. H. (1982). Cognition and language development in autism. Journal of Speech and Hearing Disorders, 47, 63-70. Wiig, E. H. (1989). Steps to language competence: Developing metalinguistic strategies. New York: The Psychological Association. Wiig, E. H., & Secord, W. (1985). Test of Language Competence. Columbus, OH: Merrill. Wiig, E. H., & Secord, W. (1989). Test of Language Competence-Expanded Edition. New York: The Psychological Corporation.

240 Wing, L. (1981). Asperger syndrome: A clinical account. Psychological Medicine, 11, 115-129. Wing, L. (1989). The diagnosis of autism. In C. Gillberg (Ed.), Diagnosis and treatment of autism (pp. 5-22). New York: Plenum Press. Wing, L. (1998). The history of Asperger syndrome. In E. Schopler, G. B. Mesibov, & L. J. Kunce (Eds.), Asperger syndrome or high-functioning autism? (pp. 11-28). New York: Plenum Press. Wolfberg, P. J., & Schuler, A. L. (2006). Promoting social reciprocity and symbolic representation in children with autism spectrum disorders. In T. Charman & W. Stone (Eds.), Social and communication development in autism spectrum disorders (pp. 180-218). London: The Guilford Press. Woodbury-Smith, M., Klin, A., & Volkmar, F. (2005). Asperger's syndrome: A comparison of clinical diagnoses and those made according to the ICD-10 and DSM-IV. Journal of Autism and Developmental Disorders, 35(2), 235-240. World Health Organization [WHO]. (1992). The ICD-10 classification of mental and behavioural disorders: Clinical descriptions and diagnostic guidelines. Geneva: Author. Zachman, L., Huisingh, R., Barrett, M., Orman, J., & LoGiudice, C. (1994). Test of Problem Solving-Elementary, Revised. East Moline, IL: LinguiSystems. Zanini, S., Bryan, K. L., De Luca, G., & Bava, A. (2005). The effects of age and education on pragmatic features of verbal communication: Evidence from the Italian version of the Right Hemisphere Language Battery (I-RHLB). Aphasiology, 19(12), 1107-1133. Zelazo, P. D., Jacques, S., Burack, J. A., & Frye, D. (2002). The relation between theory of mind and rule use: Evidence from persons with autism-spectrum disorders. Infant and Child Development, 11, 171-195.

241

242

Appendixes Appendix A

DSM-IV diagnostic criteria for Autistic Disorder

(APA, 1994, p. 70-71) A.

A total of six (or more) items from (1), (2), and (3), with at least two from (1), and one each from (2) and (3): (1)

qualitative impairment in social interaction, as manifested by at least two of the following: a) marked impairment in the use of multiple nonverbal behaviors such as eye-to-eye gaze, facial expression, body postures, and gestures to regulate social interaction b) failure to develop peer relationships appropriate to developmental level c) a lack of spontaneous seeking to share enjoyment, interests, or achievements with other people (e.g., by a lack of showing, bringing, or pointing out objects of interest) d) lack of social or emotional reciprocity

(2)

qualitative impairments in communication as manifested by at least one of the following: a) delay in, or total lack of, the development of spoken language (not accompanied by an attempt to compensate through alternative modes of communication such as gesture or mime) b) in individuals with adequate speech, marked impairment in the ability to initiate or sustain a conversation with others c) stereotyped and repetitive use of language or idiosyncratic language d) lack of varied, spontaneous make-believe play or social imitative play appropriate to developmental level Appendix continues

243

Appendix A continued (3)

restricted repetitive and stereotyped patterns of behavior, interests, and activities, as manifested by at least one of the following: a) encompassing preoccupation with one or more stereotyped and restricted patterns of interest that is abnormal either in intensity or focus b) apparently inflexible adherence to specific, nonfunctional routines or rituals c) stereotyped and repetitive motor mannerisms (e.g., hand or finger flapping or twisting, or complex whole-body movements) d) persistent preoccupation with parts of objects

B

Delays or abnormal functioning in at least one of the following areas, with onset prior to age 3 years: (1) social interaction, (2) language as used in social communication, or (3) symbolic or imaginative play.

C.

The disturbance is not better accounted for by Rett’s Disorder or Childhood Disintegrative Disorder.

244

Appendix B

DSM-IV diagnostic criteria for Asperger Disorder/syndrome

(APA, 1994, p. 77) A

Qualitative impairment in social interaction, as manifested by at least two of the following: (1)

marked impairment in the use of multiple nonverbal behaviors such as eye-to-eye gaze, facial expression, body postures, and gestures to regulate social interaction

(2)

failure to develop peer relationships appropriate to developmental level

(3)

a lack of spontaneous seeking to share enjoyment, interests, or achievements with other people (e.g., by a lack of showing, bringing, or pointing out objects of interest to other people)

(4) B.

lack of social or emotional reciprocity

Restricted repetitive and stereotyped patterns of behavior, interests, and activities, as manifested by at least one of the following: (1)

encompassing preoccupation with one or more stereotyped and restricted patterns of interest that is abnormal either in intensity or focus

(2)

apparently inflexible adherence to specific, nonfunctional routines or rituals

(3)

stereotyped and repetitive motor mannerisms (e.g., hand or finger flapping or twisting, or complex whole-body movements)

(4) C.

persistent preoccupation with parts of objects

The disturbance causes clinically significant impairment in social, occupational, or other important areas of functioning.

D.

There is no clinically significant general delay in language (e.g., single words used by age 2 years, communicative phrases used by age 3 years). Appendix continues

245

Appendix B continued E.

There is no clinically significant delay in cognitive development or in the development of age-appropriate self-help skills, adaptive behavior (other than in social interaction), and curiosity about the environment in childhood.

F.

Criteria are not met for another specific Pervasive Developmental Disorder or Schizophrenia.

246 Appendix C

Portion of questionnaire pertaining to developmental

language history COMMUNICATION HISTORY (if able to recall) • Did your child babble as an infant (e.g., “mum mum”, “doo ga”)? • At what age (approximately) did your child use recognisable words for objects or actions (e.g., “up”, “more”, “mum”, “dad”)? • At what age did your child start putting two or more words together to make little sentences (e.g., “me up”, “where mum”)? • Did you have any concerns about how your child was developing with using or understanding words? If so, please write what your concerns were. DSM-IV’s language criterion (APA, 1994) was used as the standard in determining the reclassification of ASD participants according to developmental language history. If a child’s reported onset of single words was prior to age 2 years, this was documented as Normal onset, and these children were reclassified as AS. Where the reported onset of first words was after age 2 years, this was documented as Delayed, and the child was reclassified as HFA. Where language history could not be provided, this was documented as Unsure. Inter-rater reliability for decisions made was 100 %.

247 Appendix D

Examples of the four Test of Language Competence-Expanded Edition

subtests (Wiig & Secord, 1989) Subtest

Test item example

Ambiguous Sentences

Target sentence:

An ambiguous sentence is read aloud

The elephant was ready to lift.

without the use of intonation to indicate meaning. The sentence is shown to the

Acceptable responses for full score:

participant (P). To obtain the full score,

Ready to lift something.

P must provide two different

Ready to be lifted.

interpretations.

Listening Comprehension: Making

Target prompts:

Inferences

Mother was happy to have the turkey

Two statements are read to P. The

and all the trimmings in the house.

written prompts remain visible to P. P

The family was disappointed when they

is read and shown four possible options

had to eat at a restaurant on

that link the two statements. To obtain

Thanksgiving Day.

the full score, P must select two correct

Options: (correct responses marked *)

responses.

•

the mother got sick with the flu *

•

mother forgot to buy the turkey

•

most people think that Thanksgiving dinner is always better at a restaurant

•

mother burned the turkey by cooking it too long * Appendix continues

248

Appendix D continued Subtest

Test item example

Oral Expression: Recreating Sentences

Target words:

A picture of a situation is shown to P.

Sit, painted, because.

Three target words are read out to P

Target context shown in the picture:

and remain in view of P. P is instructed

A lady is about to sit down on a park

to make up a sentence a person could

bench. She is unaware that there is a

have said using the three target words.

sign saying WET PAINT attached to

Sentences are scored on two levels. The

the bench. A young man is standing by

first level is the number of target words

the park bench looking at the lady.

used by P. The second level of scoring

Acceptable response for maximum

is an holistic approach, which examines

score:

P’s intuitions about language and

You shouldn’t sit down on that bench

communication. To obtain the

lady because it has just been painted.

maximum score, P’s sentence must be intact semantically, syntactically, and pragmatically (i.e., the sentence must be logical and meaningful, and must relate to the context presented in the picture). Appendix continues

249

Appendix D continued Subtest

Test item example

Figurative Language

Target context:

P is told of the context of an utterance

Two boys talking at a dog show.

said by someone, and the utterance

Utterance:

said. P is asked to put into words what

He is crazy about that pet.

the person meant. Next, the same

Once P has put the utterance into own

context and utterance, as well as four

words…..

possible solutions to what was meant

Options: (correct response marked *)

are presented visually to P. P selects

•

the pet makes him angry

the best interpretation for the utterance.

•

he is up in arms about the pet

P is scored for both their own

•

the pet is really wild

interpretation and the selection of the

•

he is wild about the pet *

correct interpretation.

250 Appendix E

Examples of subtest items from the Right Hemisphere Language

Battery (Bryan, 1989) Metaphor Picture Test Examiner prompt: Listen carefully to the sentence I read and point to the picture which matches the meaning of the sentence. He was very busy so she gave him a hand.

Appendix continues

251 Appendix E continued Written Metaphor Test Examiner prompt: Here is a sentence and here are some sentences which explain the meaning of this sentence. Listen to the sentence and point to the one which explains the meaning. The leader gave the group a tall order. Response choices (correct response marked *) 1. The leader gave orders to the tall group. 2. The leader measured his orders carefully. 3. The leader expected the group to do some difficult tasks. * Comprehension of Inferred Meaning Examiner prompt: I’m going to read you a paragraph and then ask you some questions about it. Listen carefully to the paragraph. (Examinee can refer to the written paragraph while answering). “I’m really hungry”, said David. “Keep an eye open on your side Sue”. They went speeding along for a few more minutes. “Look, down that side road”, she said. “Stop as soon as you can”. a. How many of them were there? (two) b. What were they riding in? (car) c. Who was driving? (David) d. What were they looking for? (somewhere to get something to eat)

Appendix continues

252 Appendix E continued Appreciation of Humour Examiner prompt: This is a joke which isn’t finished. Here are some possible punch lines. Listen to the joke and pick the punch line which makes it funny. (Printed joke remains in view of the examinee). A worried patient tells the doctor that he is still exhausted when he gets home from work. “Don’t worry,” said the doctor. “Just have a drink to wake you up”. “But last week you told me to give up drinking” said the patient. “Well,” the doctor replied: Response choices (correct response marked *) 1. “You must do as you are told”. 2. “I can’t cure everything you know”. 3. “Medical science has progressed enormously since then”. * 4. “Is it raining outside?”

Appendix continues

253 Appendix E continued Lexical Semantic Test Examiner prompt: Listen to the word I say and point to the picture which illustrates the word. •

Coat

Appendix continues

254 Appendix E continued Production of Emphatic Stress Examiner prompt: Look at these two pictures. They illustrate a sentence. I’ll start the sentence and you finish it off. (Emphatic stress is placed on the underlined word, and the expected placement of stress for the examinee’s response is underlined).

He didn’t telephone but

Expected response >

he wrote a letter

255

Appendix F

Example items from the Children’s Communication

Checklist-Second Edition (Bishop, 2003) CCC-2 Scale

Question

Question

number Coherence

10

Use terms like “he” or “it” without making it clear what s/he is talking about. For instance, when talking about a film, might say “he was really great” without explaining who “he” is.

Inappropriate

5

Initiation Stereotyped

Talks repetitively about things no-one is interested in.

11

Language

Says things that s/he does not seem to fully understand (may appear to be repeating something s/he’s heard an adult say). So, for instance, a 5-year-old may be heard to say of a teacher “she’s got a very good reputation”.

Use of Context

15

Misses the point of jokes and puns (though may be amused by nonverbal humour such as slapstick).

Nonverbal Communication

8

Looks blank in a situation where most children would show a clear facial expression – e.g. when angry, fearful or happy.

256 Appendix G

Example scenario, related questions, and scoring guide for Picture 1 of

the Test of Problem Solving-Elementary, Revised (Zachman et al., 1994) Picture 1 “Doctor”

Picture Description A young boy is being examined by a doctor. The doctor is looking in the boy’s mouth. The boy is holding a teddy bear

Question

Response must include

What’s happening in this picture?

Being with a doctor/ nurse and getting examined/checked

Suppose Martin isn’t sick, but he’s getting a

Checking for illness or wellness or

check up for preschool. What is a check-up?

two or more check-up procedures

The doctor wants to know if Martin is

Passing a disease to others or

healthy. Why is that important?

getting treatment if ill or interfering with education or living a long life

The doctor says Martin is healthy, but he

Staying healthy or preventing

needs a shot. Why?

disease or being necessary for school Appendix continues

257

Appendix G continued Question


Why does Martin have his teddy bear with

Makes him feel

him?

secure/unafraid/relaxed/not alone

258 Appendix H

Example script, related questions, and scoring guide for the Test of

Problem Solving-Adolescent (Bowers et al., 1991) Scenario K. To Date or Not to Date

Maria and Jim are teenagers who really like each other. Maria is 15 and Jim is 17, and they both go to the same high school. Jim has a job after school and on weekends, and he just bought his own car. He wants to take Maria to a movie or a party, but she isn’t allowed to date until she’s 16. It will be eight months until Maria’s birthday

Question


Why do you think Maria’s parents set the

Dating + responsibility/maturity/age or

rule of no dating until age 16?

protect her/keep her from getting hurt/in trouble or parent’s dating rules when they were young

Do you think the rule of no dating until age

Irresponsible behavior/immaturity or

16 is fair? Why?

responsible behavior/maturity or keeping out of trouble/protected

What kind of person do you think Jim is?

Jim working + saving money

How can you tell?

Maria’s parents don’t like the fact that

Jim + his age/reputation/trustworthiness or

Maria is interested in Jim. Why might they

Maria + trustworthiness

feel this way? Appendix continues

259

Appendix H continued Question


Jim is willing to follow Maria’s parents’

Respecting her parents/their reasons or

rule. Why?

liking/wanting to date/protecting + Maria

What activities might Maria’s parents

A reasonable group/supervised/school

allow Maria and Jim to do now?

activity or a reasonable activity in a public place/at home

Why would Maria’s parents let them do

Supervision/group participation/occurring

these activities?

in a public place

260 Appendix I

Rating Scale of Pragmatic Communication Skills

(Burns et al., 1985) Rating

1

2

3

4

5

A. Nonverbal Communication Intonation

Flat or stereotyped

Limited or inappropriate

Appropriate

Facial Expression

None

Limited or inappropriate

Appropriate

Eye Contact

Cannot establish or maintain eye contact

Needs cues to establish or maintain eye contact

Appropriate

Gestures and Proxemics

Inappropriate or does not use

Inconsistent appropriate use

Appropriate

B. Conversational Skills Conversational Initiation

Inappropriate or does not initiate

Inconsistent appropriate initiation

Appropriate

Turn-taking

Does not obey signals

Inconsistently responsive to signals

Adequate

Verbosity

Over 50 % of responses are verbose or tangential

25 % to 50 % of responses are verbose or tangential

Appropriate response length Appendix continues

261

Appendix I continued Rating

1

2

3

4

5

C. Use of Linguistic Context Topic Maintenance

Maintains topic less than 25 % of the time

Maintains topic 50 % of the time

Maintains topic

Presupposition

Presupposes too much and/or too little 50 %

Presupposes too much and/or too little 25 %50 %

Appropriate

Referencing Skills

Inappropriate referencing

Inconsistent appropriate referencing

Appropriate

Organisation of a Narrative Organisation

Disorganised

Some organisation but lacks a unifying theme

Adequate

Completeness

More than 50 % of details are missing and/or inaccurate

25 %-50 % of details are missing and/or inaccurate

Adequate

262

Appendix J

Assessments administered to all adult participants in the gender

study Assessment

Details

Test of Nonverbal

TONI-2 is a language-free measure of cognitive

Intelligence-Second Edition

ability, and was included in the assessment to

(TONI-2) (Brown et al.,

determine whether verbal performance reflected

1990)

nonverbal performance. Raw scores were converted to Standard Scores as per test manual.

Western Aphasia Battery

The WAB was chosen to assess linguistic function.

(WAB) (Kertesz, 1982)

For the purposes of the study, only the language component of the WAB was administered. The subtests administered were Spontaneous Speech (Information and Fluency), Auditory Verbal Comprehension, Repetition, and Naming. These subtests provide an Aphasia Quotient (AQ). The WAB AQ was used in the analyses. The WAB is standardised for the adult population, and is intended to be used for adults with brain damage. Appendix continues

263

Appendix J continued Assessment

Details

Right Hemisphere Language

The RHLB was chosen to assess pragmatic language

Battery (RHLB) (Bryan,

skills. The RHLB consists of seven subtests, six of

1989)

which were used in the study. The subtests included were Metaphor Picture Test, Written Metaphor Test, Comprehension of Inferred Meaning, Appreciation of Humour, Lexical Semantic Test, and Production of Emphatic Stress. Adult raw scores were converted to relative T scores as outlined in the test manual.

Test of Problem Solving-

Social problem solving was assessed using the

Adolescent (TOPS-A)

TOPS-A. TOPS-A is not standardised for the adult

(Bowers et al., 1991)

population. Raw scores were used for the statistical analyses.

Test of Language

Level 2 of TLC-E was administered. The TLC-E

Competence-Expanded

assesses metalinguistic competence in semantics,

Edition (TLC-E) (Wiig &

syntax, and/or pragmatics. The TLC-E is not

Secord, 1989)

standardised for the adult population; therefore raw scores were used for the statistical analyses. The TLC-E subtests are Ambiguous Sentences, Listening Comprehension: Making Inferences, Oral Expression: Recreating Sentences, and Figurative Language. Appendix continues

264

Appendix J continued Assessment

Details

Rating Scale of Pragmatic

Subjective assessment of social pragmatic skills was

Communication Skills

provided by the RSPCS. Each participant completed

(RSPCS) (Burns et al., 1985)

a self-rating, and a person familiar with the participant completed a Significant Other rating. The RSPCS has four sections: Nonverbal Communication: use of intonation, facial expression, eye contact, and gestures and body proxemics; Conversational Skills: conversation initiation, turntaking, and verbosity; Use of Linguistic Context: topic maintenance, presupposition, and referencing skills; Organisation of a Narrative: organisation and completeness of discourse.