Speeding Personality Measures to Reduce Faking

9 downloads 0 Views 255KB Size Report
Wilfrid Laurier University, Waterloo, ON, Canada. Abstract. .... Barksdale, Robin, & James, 2007), forced-choice measures .... Wiechmann & Ryan (2003).
Original Article

Speeding Personality Measures to Reduce Faking A Self-Regulatory Model Shawn Komar,1 Jennifer A. Komar,1 Chet Robie,2 and Simon Taggar2 1

Department of Psychology, University of Waterloo, ON, Canada, 2School of Business & Economics, Wilfrid Laurier University, Waterloo, ON, Canada Abstract. The purpose of the present study was to examine the effects of imposing a time constraint on respondents completing the Big Five personality Inventory (John & Srivastava, 1999) based on a self-regulatory model of response distortion. A completely crossed 2 · 2 experimental design was used in which instructions (neutral standard instruction or a job applicant instruction) and speed (with or without a time limit) were manipulated. While speeding personality tests reduced socially desirable responding, consistent with resource allocation theory (Ackerman, 1986), this effect was only seen in low cognitive ability individuals. Speeding was not perceived negatively by participants. This study is the first to find any evidence of a possible influence of speed on impression management and suggests that manipulating time limits for completing personality measures in selection is not advised at the present time as it is likely to have the unintended effect of removing applicants with high cognitive ability from the applicant pool. Keywords: Big Five, personality, impression management, speeding, faking

Personality testing is a $400 million industry in the US alone and is growing at a 10% annual rate (Hsu, 2004). In a survey of 959 organizations from 20 different countries, personality inventories were found to be the most often used method of selection (Ryan, McFarland, Baron, & Page, 1999). There appears to be good reason for the escalating popularity of personality tests. Barrick, Mount, and Judge’s (2001) quantitative summary of 15 meta-analytic studies found support linking the Five Factor Model personality traits to job performance. Nonetheless, concerns about the use of personality measures for personnel selection persist (e.g., Murphy & Dzieweczynski, 2005). Robie, Tuzinski, and Bly (2006) found that 70% of 77 experienced US assessors believed faking was a serious threat to the validity of the personality inventory in the assessment process and that one-quarter of applicants were expected to fake on the personality inventory. Further, studies have also found that faking can affect hiring decisions (Rosse, Stecher, Miller, & Levin, 1998) and criterion-related validity (Komar, Brown, Komar, & Robie, 2008). According to resource allocation theory (Ackerman, 1986; Kanfer & Ackerman, 1989), imposing time restrictions on the completion of paper-and-pencil personality measures, referred to as speeding, may represent one way of reducing faking. The theory proposes that individuals possess a finite amount of attentional resources that can be 1

directed toward one or multiple activities. It suggests that when respondents are placed under a time pressure they may be less able to fashion faked responses. However, the self-regulatory mechanisms described in resource allocation theory have been criticized. For example, DeShon, Brown, and Greenis (1996) argued that self-regulation does not require that substantial attentional resources be diverted from task performance. Accordingly, we test the efficacy of a self-regulatory model of faking personality tests.

The Faking Problem Faking involves impression management (IM; Goffman, 1959), where the personality test respondent tries to control the impression he or she makes. It is an applicant’s goaldirected conscious or unconscious attempt to influence the organization’s perceptions by regulating and controlling information. Faking is typically considered a trait-irrelevant variance that degrades decisions that are based on trait scores (Zickar & Robie, 1999).1 Practitioners often include IM scales, also known as social desirability scales, in assessment batteries to identify fakers and correct personality scores for faking (Robie et al., 2006). IM captures the favorable representation of the self to others or ‘‘a positive public impression’’ (Paulhus,

Faking may actually not be wholly trait-irrelevant for some jobs (e.g., sales), although this is likely the exception and not the rule (Komar et al., 2008; Tull, 1998).

Journal of Personnel Psychology 2010; Vol. 9(3):126–137 DOI: 10.1027/1866-5888/a000016

 2010 Hogrefe Publishing

S. Komar et al.: Speeding Personality Measures and Faking

2001). These scales typically consist of items written to reflect desirable behaviors that are very rare or undesirable behaviors that are relatively common. Respondents who indicate a high level of agreement with the desirable behaviors and who eschew the undesirable behaviors are thought to be impression managing by using a socially desirable response set. Although widely used, researchers have found limited support for IM scales (Burns & Christiansen, 2006) and some have questioned the assumption that impression managers tend to endorse the extreme response options on such inventories, calling into question the typical scoring and use of these measures (Kuncel & Tellegen, 2009). Past efforts at preventing faking on personality inventories, such as warnings (Dwight & Donovan, 2003), subtle items used in conditional reasoning measures (LeBreton, Barksdale, Robin, & James, 2007), forced-choice measures (Jackson, Wroblewski, & Ashton, 2000), or the knowledge response format of situational judgment tests (Hooper, Cullen, & Sackett, 2006), have not been met with a great deal of success. Reactive approaches aimed at identifying individuals who likely faked on a personality inventory have also yielded limited utility. For instance, a preponderance of evidence suggests that self-report measures of social desirability cannot profitability be used to remediate faking (Ones, Viswesvaran, & Reiss, 1996; Schmitt & Oswald, 2006). Also, the use of item response theory (IRT) (i.e., appropriateness measurement) to identify fakers has not improved identification over that of self-report social desirability measures (Robie, Zickar, & Schmit, 2001), although mixed-model IRT, an emerging statistical technique, has shown some success in differentiating individuals who are extreme fakers from those who are only slightly faking or who are answering honestly (cf. Zickar, Gibby, & Robie, 2004). Researchers have had some success using response latencies to detect fakers on personality measures in a personnel selection context (Holden, 1995, 1998; Holden & Hibbs, 1995; Vasilopoulos, Reilly, & Leaman, 2000). Research has shown that IM leads to faster responding on self-report measures under certain conditions, for example when respondents are highly familiar with the job (Vasilopoulos et al., 2000). Other research suggests that it takes respondents longer to formulate a faked response. Response times are longer when the respondent is unfamiliar with the job or is providing a response that is inconsistent with the self-schema (Holden, 1995), as when one is trying to fake good. Robie, Brown, and Beaty (2007) found evidence that individuals who faked took longer to complete a personality inventory on average than those answering honestly. These results are the impetus for the present study. That is, we assess the impact of imposing a time constraint (i.e., speeding) on faking when applicants complete a personality test.

The Impact of Speeding A self-regulatory model of faking. A basic premise of resource allocation theory (Ackerman, 1986) is that individ2

127

uals have a limited amount of attention and effort that can be allocated to a range of different activities, including on-task, off-task, and self-regulation (such as, active monitoring of the impression one creates) activities (Kanfer & Ackerman, 1989). Self-regulatory cognitions would be initiated when the familiar activity of disclosing truthful behavioral predispositions, or habitual patterns, is impeded and the test taker’s goal becomes to impression manage. With respect to personality measurement, resource allocation theory may help to explain unexpected empirical findings. For instance, Robie et al. (2000) found that respondents were able to ‘‘beat’’ response latency measurements (i.e., avoid taking too long to respond to a personality question) when they were coached on how to do so. However, they were not able to elevate their personality scores on average in comparison to a group told to answer honestly. That is, they were unable to fake. In explaining these results, Robie et al. (2000) suggested that participants in the coached condition may have lacked the attentional resources to simultaneously control response time and respond desirably to items. The use of speeded tests has a long history in cognitive ability testing (Aitkens, Thorndike, & Hubbell, 1902; Evans & Reilly, 1972; Freeman, 1928)2 and has been used successfully to reduce cheating on cognitive ability tests (Arthur, Glaze, Villado, & Taylor, 2009). Arthur and colleagues argue that imposing a time limit on completing a cognitive ability test in an unproctored setting inhibits test takers’ ability to use aids or obtain help from others. Although posited to operate through different means, speeding is a promising method for eliciting honest responding on both cognitive ability tests and personality scales. However, it is surprising that speeding has received almost no attention as a preventative approach to faking on personality tests. Thus, resource allocation theory (Ackerman, 1986) and some empirical findings suggest that time constraints reduce respondents’ ability to allocate cognitive resources to the self-regulation required to faking-good on a personality inventory. We predict an interaction between faking and speeding such that participants motivated to fake will score better on the Big Five personality and IM scales compared to those told to respond honestly, but only when given ample time to complete the scales. Specifically: Hypothesis 1a: When instructed to respond as an applicant, respondents to a Big Five personality measure will evidence statistically higher personality scale scores for extraversion, agreeableness, conscientiousness, and openness in unspeeded conditions than those of participants in speeded conditions. Hypothesis 1b: When instructed to respond as an applicant, respondents to a Big Five personality measure will evidence statistically lower Neuroticism scale scores in unspeeded conditions than those of participants in speeded conditions.

An annotated bibliography reviewed 63 studies of test speed in cognitive testing conducted from the years 1902–1976 (Donlon, 1980).

 2010 Hogrefe Publishing

Journal of Personnel Psychology 2010; Vol. 9(3):126–137

128

S. Komar et al.: Speeding Personality Measures and Faking

Hypothesis 2: When instructed to respond as an applicant, respondents in unspeeded conditions will evidence statistically lower IM scores than those of participants in speeded conditions.

The Role of Cognitive Ability in Self-Regulation During Speeding One particularly practical question is how the relationship between cognitive ability and IM is affected by job applicants’ goals to fake-good under time constraints. Cognitive ability and social desirability have a negligible correlation with one another under normal circumstances (Ones et al., 1996), although several studies have shown a positive relationship between cognitive ability and faking ability (e.g., Brown & Cothern, 2002; Pauls & Crost, 2005). According to cognitive load theory, individuals with higher cognitive ability are better able to handle tasks requiring higher cognitive load (Sweller, Van Merrie¨nboer, & Paas, 1998) and are less likely to suffer from the depletion of attentional resources than are less intelligent individuals. Thus, individuals with relatively high general cognitive ability (g) should be better equipped to handle any higher attentional demands when asked to fake under time constraints. In addition to the attentional resources needed to read and comprehend the items and response options on a personality test, a test-taker engaging in socially desirable responding must also evaluate the impression they are creating as they answer each item. Further self-regulation is required when completing the questionnaire under a time limit, as participants must also control how much time they spend answering each item. These self-regulatory activities consume valuable attentional resources that may be better applied to task performance and thus may result in lower task performance (Kanfer & Ackerman, 1989). If high g individuals are better equipped to provide such resources, then high g individuals who are motivated to impression manage will score lower on IM under speeded conditions than low g individuals. In the unspeeded fake-good condition, cognitive ability is unlikely to have an impact on IM as dealing with a high cognitive load should be less of an issue.

Perceived Procedure Characteristics

Change to Personality Test

Attitudes Toward Speeding Personality Tests Negative reactions toward personality tests are a particularly important concern because they generally possess less face validity than other personnel selection tools (Ambrose & Rosse, 2003). Neutral (to negative) reactions are related to important applicant attitudes and intentions such as: the likelihood of viewing the organization favorably and reporting stronger intentions to accept job offers and recommending the employer to others (Hausknecht, Day, & Thomas, 2004); applicants’ motivation to do well (Arvey, Strickland, Drauden, & Martin, 1990); withdrawal from the applicant pool (Rynes, Bretz, & Gerhart, 1991); applicants’ self-efficacy (Ployhart & Ryan, 1997); lawsuits (Bible, 1990); and loyalty to the employer after having been hired (Crant & Bateman, 1990). Thus, it is important to examine the effect of applicant reactions on any changes to typical personality testing (e.g., adding a speed component) in order to mitigate possible negative consequences. Although more research is needed to link these perceptions to applicant behavior (Sackett & Lievens, 2008), it is important to demonstrate whether different aspects of the selection system affect applicant reactions. Figure 1 portrays a model of how speeding personality tests may affect applicant reactions. We proposed that speeding a personality test would decrease: (a) face validity because applicants will not perceive a strong link between the speed of responding to personality items and job performance; (b) opportunity to perform because applicants will not have as much time to fully contemplate their answers and present themselves in the ways in which they would prefer to be seen; and (c) test ease because finishing in the allotted time will place a cognitive and emotional strain

Applicant Perceptions

+

Face validity

Motivation

-

-

Hypothesis 3: There will be three-way interaction between the speeding condition, faking condition, and respondent g, such that IM scores will be highest for low g individuals who were given the goal to fakegood in 10 min. Conversely, high g respondents in the speeded fake-good condition will score relatively low in IM.

+ Personality test speeding

Figure 1. Hypothesized model of effects of personality test speeding on perceived procedure characteristics and applicant perceptions.

Self-Doubts Regarding TestTaking

Opportunity to perform -

Dislike of Tests Perceived test ease

-

Non-validated relationship Hausknecht et al. (2004) Wiechmann & Ryan (2003)

Journal of Personnel Psychology 2010; Vol. 9(3):126–137

 2010 Hogrefe Publishing

S. Komar et al.: Speeding Personality Measures and Faking

on participants. Research by Hausknecht et al. (2004) and Wiechmann and Ryan (2003) suggests that decreases in: (a) face validity will be related to decreases in motivation and increases in dislike for tests; (b) opportunity to perform will be related to decreases in motivation; and (c) perceived test ease will be related to increases in self-doubts regarding test-taking. Based on this model, we proposed the following set of related hypotheses: Hypothesis 4(a–c): Participants in the speeded conditions in comparison to participants in the unspeeded conditions will experience significantly lower motivation (Hypothesis 4a), higher self-doubts regarding test-taking (Hypothesis 4b), and higher dislike of tests (Hypothesis 4c).

Method Participants Participants were 243 undergraduate students (57.6% female and 71.6% White) enrolled at a comprehensive North American university who participated for extra course credit. Mean age was 20.74 (SD = 2.44); one participant did not report their age. Students who were enrolled in one of several Organizational Behavior and Human Resource Management courses were eligible to participate. Most of the students were Business or Economics majors (87.2%) with the remainder being Sciences (i.e., Psychology) or Arts (i.e., Communications, Geography, Global Studies, History, and Sociology) majors. No statistically significant differences were observed across experimental cells on gender, race, or major.

Measures Personality The Big Five Inventory (BFI; John & Srivastava, 1999) contains 44 short, phrase-based items which measure the five major components of personality including: extraversion, agreeableness, conscientiousness, neuroticism, and openness. Items were anchored by a 5-point Likert-type scale (disagree strongly to agree strongly). Impression Management We employed a measure of IM developed by researchers involved in the International Personality Item Pool (IPIP; Goldberg et al., 2006). The IPIP IM scale was designed to parallel Paulhus’s (1991) IM scale from the Balanced Inventory of Desirable Responding (BIDR), which is widely used by researchers studying IM. The BIDR has moderate to high correlations with other measures of social desirability that are commonly used both in research and practice (Lanyon & Carle, 2007; Paulhus, 1991). The IPIP IM scale has been found to correlate .79 (.96 when corrected for attenuation)  2010 Hogrefe Publishing

129

with the BIDR (Goldberg et al., 2006). Our IM scale contained 20 items, as did the IM BIDR scale, and exceeded the reliability (a = .88) of any of those reliabilities reported by Paulhus (1991). An example item on the IPIP IM scale of a desirable but rarely performed behavior is, ‘‘Always admit it when I make a mistake,’’ (positively keyed) and an example of an undesirable but common behavior was, ‘‘Use swear words’’ (negatively keyed). Test Attitude Survey The Test Attitude Survey (TAS; Arvey et al., 1990) is a 45-item instrument that assesses attitudes toward the personality measure and is rated on a 5-point Likert-type scale (1 = disagree strongly to 5 = agree strongly). McCarthy and Goffin (2003) found that the TAS was best interpreted in terms of three factors: (1) Motivation (14 items; e.g., ‘‘I tried to do the very best I could on this test’’); (2) selfdoubts regarding test-taking (11 items; e.g., ‘‘During the testing, I often thought about how poorly I was doing’’); and (3) Dislike of tests (15 items; e.g., ‘‘I don’t believe that tests are valid’’). General Cognitive Ability The Wonderlic Personnel Test (WPT) is a 50-question exam to assess aptitude for learning a job and adapting to solve problems for employees in a wide range of occupations (Wonderlic, 1992). The WPT yields one final score which is the sum of correct answers. The average score on the WPT for all US job applicants and high school graduates is 21 and the average score for US college graduates is 29. The average WPT score for our sample was 24.12 (SD = 4.79). Manipulation Checks Participants in all conditions were asked five manipulation check questions embedded in the TAS. An example item was, ‘‘I felt rushed when completing the test.’’ We averaged the items in the scale as it evidenced high reliability (a = .86) and a unifactor solution (first component in a PCA accounted for 65% of the variance). A large and statistically significant mean difference was evidenced between the two groups such that individuals assigned to the speeded conditions (N = 123) reported more strongly that they felt rushed and did not have enough time to complete the test (M = 3.51, SD = 0.87) than did individuals assigned to nonspeeded conditions (N = 120) (M = 1.63, 0.68) (t = 18.72, p < .001, Cohen’s d = 2.41). Participants in the faking conditions (N = 126) were asked several questions embedded in the demographic questionnaire regarding their perceptions of the incentive used in the study. Responses were anchored on a 5-point Likert-type scale (disagree strongly to agree strongly). The average response to the first question, ‘‘I would like to receive the $20 incentive’’ was 4.66 (SD = 0.65). The average response to the second question, ‘‘I need the $20 incentive’’ was 2.52 Journal of Personnel Psychology 2010; Vol. 9(3):126–137

130

S. Komar et al.: Speeding Personality Measures and Faking

(SD = 1.49). The average response to the third question, ‘‘Winning the $20 incentive is NOT important to me’’ was 2.60 (SD = 1.23). It appears from the responses to these questions that participants overwhelmingly wanted to receive the $20 but that there was variability in the degree to which the money was needed or important to participants.

Design and Procedure Participants were randomly assigned in groups of 5–30 individuals to one of four experimental cells: (1) honest-unspeeded; (2) honest-speeded; (3) faked-unspeeded; and (4) faked-speeded. Students were scheduled for testing using a web-based human subject pool management software package for universities. Students signed up for timeslots on a first-come, first-served basis. In the honest conditions, participants were instructed to respond honestly and accurately, and were assured that their responses would be anonymous. Participants in the faking conditions were given a job advertisement3 to read and told to answer the questions on the personality test in a way that would get them hired for the job. Additionally, an incentive of $20 was offered for those who scored in the top 15%. In the unspeeded conditions, participants were told that there was no time limit for completing the personality test, and those in the speeded conditions were told that they had 10 min to complete the test. The full instructions for each condition are presented in Appendix A. Participants first completed the consent form, followed by the personality and IM measures, the attitude measure, the cognitive ability measure, and lastly a demographic questionnaire and manipulation check measure. Participation was concluded after a debriefing statement was circulated to participants. The first 64 items on the first page of the personality test (44 personality items and 20 IM items) were scored. The remaining 52 items on the second page of the personality test contained parallel items from the IPIP and were not

3 4 5

6

scored.4 Pretesting with 13 individuals found that the first 64 items were completed under normal conditions (i.e., answer honestly, unspeeded) in an average of 5.29 min (SD = 1.83, range 2.5–8 min) and that the entire personality test was completed in an average of 10.58 min (SD = 2.60, range 6–15 min). We thus believed from this pretesting that: (a) all participants would have enough time to complete the first 64 items of the personality test; and (b) there would in all likelihood be a perception of speeding if a restriction of 10 min were placed on completing the entire personality test. Three individuals in the speeded experimental conditions did not complete the first 64 items of the personality test (1 in the faking condition and 2 from the honest condition). Missing values were replaced with means from the non-missing items. Over half of the individuals in the speeded experimental cells (63 individuals) did not complete the entire personality test. All unspeeded participants completed the entire personality test.

Results Table 1 shows the descriptive statistics and intercorrelations for all relevant study variables. Prior research has suggested that faking can impact the factor structure of personality (Ellingson, Smith, & Sackett, 2001). Analyses revealed that this was also true for the data in this study5, and so we include separate statistics and intercorrelations for individuals in the honest (Table 2a) and faked (Table 2b) conditions. Speeding did not affect the factor structure of personality6. Examination of the results in Table 3 shows that there was no statistically significant interaction between the faking and speeding manipulations on the personality or IM variables. Thus, no support was found for the first or second hypotheses. The third hypothesis stated that the relationship between cognitive ability and IM would be negative when

The job advertisement was adapted from one on the Human Resources Development Canada Job Bank web page (http://jb-ge.hrdcdrhc.gc.ca/) with identifying information on the employer and address made fictitious. We reproduce this job advertisement in Appendix B. We designed the personality inventory in such a way that all items could be completed by all participants to avoid some of the more obvious psychometric problems associated with speeded measures (cf. Cronbach & Warrington, 1951). In order to compare the factor structure of personality for faked and honest conditions, we used structural equation modeling to conduct a confirmatory factor analysis concurrently for both the fake-good and honest participants. When factor loadings, variance of latent factors, and their covariances were set equal between the two groups the Chi-square for the resulting model was 3,291.9, df = 1,838, p < .01 and the AIC was 3751.9. Relaxing the equality in factor loadings (but retaining equality for the variance and covariance of latent factors) resulted in a Chi-square of 3182.8, df = 1,799, p < .01 and an AIC of 3,720.8. A test of the difference in Chi-square suggested a significant improvement in fit (Chi-square difference of 109.1, df = 39, p < .01), a finding further supported by the smaller AIC for this model. Finally, we relaxed the equality on variances and covariances of the latent personality factors. The resulting model had a Chi-square of 3049.6, df = 1,784, p < .01 and an AIC of 3617.6, which represented a significant improvement over the model which allowed factor loadings to differ (i.e., Chi-square difference of 133.2, df = 15, p < .01); the factor structure of personality was affected by faking. We used structural equation modeling to conduct a confirmatory factor analysis concurrently for both the speeded and unspeeded conditions. As a first step, we held factor loadings, variance of latent factors, and their covariances, equal between the two groups. This model had a Chi-square of 3004.4, df = 1,838, p < .01 and an AIC of 3464.4. We then allowed the factor loadings to differ between the two groups but retained the restrictions on variances and covariances of the latent factors. The Chi-square of the resulting model had a Chi-square of 2975.1, df = 1,799, p < .01 and an AIC of 3513.01. Note that both a comparison of the AIC and a statistical test of the Chi-square difference (i.e., Chi-square difference of 29.3, df = 39, p = 0.87) suggested retaining the model with greater parsimony; there was no difference in the factor structure of personality between the two speeding conditions.

Journal of Personnel Psychology 2010; Vol. 9(3):126–137

 2010 Hogrefe Publishing

S. Komar et al.: Speeding Personality Measures and Faking

131

Table 1. Descriptive statistics and intercorrelations across conditions Variable 1. GCA 2. Fake 3. Speed 4. E 5. A 6. C 7. N 8. O 9. IM 10. TAS1 11. TAS2 12. TAS3

M

SD

24.12 0.52 0.51 3.86 4.16 4.29 2.27 3.86 3.74 3.71 2.13 2.25

4.79 0.50 0.50 0.75 0.61 0.63 0.87 0.64 0.62 0.65 0.54 0.64

1

2

3

4

5

6

7

8

9

10

11

12

– .02 .45 .50 .56 .49 .45 .61 .41 .00 .22

– .04 .03 .03 .01 .03 .06 .24 .43 .11

.86 .34 .45 .46 .53 .34 .27 .05 .17

.83 .61 .56 .43 .68 .34 .13 .32

.85 .63 .44 .69 .44 .18 .43

.89 .43 .57 .24 .28 .23

.80 .45 .30 .08 .28

.88 .39 .02 .29

.88 .10 .50

.72 .32

.85

a

– .08 .06 .03 .12 .00 .09 .05 .02 .04 .03 .06

Note. N = 243. Correlations  |.13| are significant at the .05 alpha level. Correlations  |.17| are significant at the .01 alpha level. Correlations  |.22| are significant at the .001 alpha level. Reliability coefficients are arranged along the diagonal. GCA = general cognitive ability. Fake (0 = honest, 1 = faked). Speed (0 = speeded, 1 = unspeeded). E = Extraversion. A = Agreeableness. C = Conscientiousness. N = Neuroticism. O = Openness. IM = impression management. TAS1 = motivation. TAS2 = self-doubts regarding test-taking. TAS3 = dislike of tests. a The WPT manual (Wonderlic, 1992) reports test-retest coefficients of .82–.94.

Table 2. Descriptive statistics and intercorrelations for the (a) honest and (b) faking conditions Condition

Variable

M

SD

1

2

3

4

5

6

7

8

(a) Honest 1. 2. 3. 4. 5. 6. 7. 8.

GCA Speed E A C N O IM

24.49 0.50 3.51 3.85 3.92 2.72 3.57 3.35

4.94 0.50 0.77 0.61 0.63 0.83 0.61 0.54

–a .09 .08 .10 .02 .15 .08 .01

– .04 .03 .01 .05 .00 .08

.87 .03 .13 .13 .30 .12

.81 .43 .50 .22 .59

.81 .38 .10 .47

.87 .12 .37

.76 .11

.83

1. 2. 3. 4. 5. 6. 7. 8.

GCA Speed E A C N O IM

23.77 0.52 4.19 4.45 4.63 1.86 4.14 4.11

4.64 0.50 0.58 0.43 0.41 0.70 0.53 0.45

–a .04 .11 .08 .16 .15 .28 .09

– .03 .02 .07 .05 .05 .03

.77 .34 .54 .58 .58 .42

.74 .52 .32 .34 .48

.76 .69 .53 .64

.86 .48 .41

.75 .44

.79

(b) Faking

Note. In the honest condition: N = 117. Correlations  |.16| are significant at the .05 alpha level. Correlations  |.22| are significant at the .01 alpha level. Correlations  |.29| are significant at the .001 alpha level. In the faking condition: N = 126. Correlations  |.15| are significant at the .05 alpha level. Correlations  |.21| are significant at the .01 alpha level. Correlations  |.28| are significant at the .001 alpha level. Reliability coefficients are arranged along the diagonal. GCA = general cognitive ability. Fake (0 = honest, 1 = faked). Speed (0 = speeded, 1 = unspeeded). E = Extraversion. A = Agreeableness. C = Conscientiousness. N = Neuroticism. O = Openness. IM = impression management. a The WPT manual (Wonderlic, 1992) reports test-retest coefficients of .82–.94.

individuals are instructed to fake a personality inventory under a speeded condition. As shown in Table 4, a three-way interaction was evidenced between cognitive ability, faking condition, and speed condition in predicting IM ratings (DR2 = .013, p < .05). Figure 2 shows the nature of this interaction. The slope for the cognitive ability 2010 Hogrefe Publishing

IM relationship was negative for participants in the fakedunspeeded condition and this slope was positive for participants in the faked speeded conditions; these slopes were significantly different from one another (t = 2.38, p < .05). These results are contrary to our third hypothesis. Journal of Personnel Psychology 2010; Vol. 9(3):126–137

132

S. Komar et al.: Speeding Personality Measures and Faking

Table 3. Regression summaries for personality and IM Extraversion

Predictor

B

Intercept Faking Speeding Faking · Speeding

3.479 .691 .064 .030

(0.088) (0.124)** (0.125) (0.174)

Intercept Faking Speeding Faking · Speeding

3.834 .610 .034 .018

(0.069) (0.097)** (0.098) (0.136)

Intercept Faking Speeding Faking · Speeding

3.932 .669 .016 .073

(0.069) (0.096)** (0.098) (0.135)

Intercept Faking Speeding Faking · Speeding

2.676 .778 .086 .155

(0.100) (0.140)** (0.142) (0.197)

Openness

Intercept Faking Speeding Faking · Speeding

3.571 .535 .002 .059

(0.075) (0.105)** (0.106) (0.147)

IM

Intercept Faking Speeding Faking · Speeding

3.309 .691 .085 .055

(0.065) (0.091)** (0.092) (0.128)

Agreeableness

Conscientiousness

Neuroticism

Note. N = 243. **p < .001. Standard errors appear in parentheses.

4.4

Impression management

Dependent variable

4.5

4.3 4.2 4.1 4.0 3.9 Honest Unspeeded Honest Speeded Faked Unspeeded Faked Speeded

3.8 3.7 3.6 3.5 3.4 3.3 3.2 3.1

-1 SD Below mean

+1 SD Above mean

General cognitive ability

Figure 2. Self-rating ratings of IM regressed on general cognitive ability (GCA) scores, faking (0 = honest, 1 = faked), and speeding (0 = speeded, 1 = unspeeded). Only scores ±1SD from the mean of self-ratings of GCA scores are plotted. The following are the slope difference tests (Dawson & Richter, 2006): Pair of slopes

t-value for slope difference

p-value for slope difference

HS vs. HU

0.69

.49

HS vs. FS

2.18

.03

HS vs. FU

0.10

.92

HU vs. FU

0.84

.40

HU vs. FS

1.66

.10

FS vs. FU

2.38

.02

HS = honest-speeded. HU = honest-unspeeded. FS = faked-speeded. FU = faked-unspeeded.

Our fourth hypothesis stated that participants in the speeded conditions, in comparison to participants in the unspeeded conditions, would experience significantly lower motivation (H4a), higher self-doubts regarding test-taking (H4b), and higher dislike of tests (H4c). As can be seen in Table 5, significant effects (p < .001) for speed were found for motivation and self-doubts regarding test-taking such that participants in speeded condition reported being more motivated and experiencing more self-doubts regarding test-taking. There was no significant association between speeding and disliking tests. These results fail to support H4a and H4c, but do support H4b.

Discussion We presented a series of hypotheses (H1–H3) in this study based primarily on the supposition that placing a restriction on the time to complete a personality measure would reduce possible faking on that measure. We did not find an effect for speeding on either personality or IM scores (H1–H2). We did find a three-way interaction between the speeding condition, faking condition, and respondent g, such that IM scores would be highest for high g individuals who were given the goal to fake-good in 10 min. Although this three-way interaction did not take the hypothesized form, we believe it is still consistent with the Journal of Personnel Psychology 2010; Vol. 9(3):126–137

The hypothesized slope difference test is underlined.

theory and with the faking strategy that participants were likely using. Participants in the faking conditions likely believed that the best strategy for responding to the IM scale was to appear extremely virtuous. Although this strategy would result in participants being tagged as a faker, they were unlikely aware of this scale’s true use. This rationale suggests that those with higher g were better able to handle the increased cognitive load imposed by faking under time pressure. Support was found for only one part of the three-part fourth hypothesis (H4b). Significant effects for speed were found such that participants in speeded condition reported being more motivated and experiencing more self-doubts regarding test-taking. Although, it should be noted that even in the speeded conditions, ratings of self-doubt were below the scale midpoint. In addition, speeding did not lead to greater dislike of tests. These results suggest that speeding does not negatively affect applicant reactions and may have positive effects on motivation.

Theoretical Implications As noted above, we did not find a main effect for speeding in the current study. An alternative theoretical mechanism may  2010 Hogrefe Publishing

S. Komar et al.: Speeding Personality Measures and Faking

133

Table 4. Regression results for IM Variable

B: Step 1

Step 1: GCA Fake Speed Step 2: GCA · Fake GCA · Speed Fake · Speed Step 3: GCA · Fake · Speed R2 DR2 Constant

B: Step 2

0.003 (0.007) 0.756 (0.064)** 0.054 (0.064)

0.370** 0.370** 3.323

B: Step 3

0.008 (0.011) 0.782 (0.091)** 0.080 (0.093)

0.004 (0.012) 0.775 (0.091)** 0.092 (0.092)

0.010 (0.013) 0.015 (0.014) 0.047 (0.129)

0.015 (0.018) 0.014 (0.019) 0.050 (0.128)

0.376** 0.005 3.309

0.059 (0.027)* 0.388** 0.013* 3.310

Note. N = 243. *p < .05. **p < .001. Standard errors appear in parentheses.

Table 5. Regression summaries for effects of speeding on TAS outcome variables Dependent variable TAS1 TAS2 TAS3

Intercept 3.559 1.893 2.323

B 0.306 (0.081)** 0.463 (0.063)** 0.139 (0.082)**

Note. N = 243. **p < .001. Standard errors appear in parentheses. TAS1 = motivation. TAS2 = self-doubts regarding test-taking. TAS3 = dislike of tests.

have accounted for these results. Some previous research suggests that self-regulation does not necessarily require high amounts of attentional resources. Control theory and several supporting studies (Bargh, 1989; DeShon et al., 1996; Payne, Bettman, & Johnson, 1993) suggest that the process of selfregulation can be highly automated through extensive practice in everyday life and thus does not deplete significant amounts of attentional resources (Lord & Levy, 1994). DeShon et al. (1996) conclude that self-regulation ‘‘can often be performed automatically and without significant cognitive resource requirements’’ (p. 597). Additional indirect evidence from this study also supports the conclusion that speeding on personality tests is not cognitively demanding. Research by Vasilopoulos, Cucina, and McElreath (2005) and Vasilopoulos, Cucina, Dyomina, Morewitz, and Reilly (2006) suggests that cognitive ability and personality, which are normally not related (e.g., McHenry, Hough, Toquam, Hanson, & Ashworth, 1990), are significantly correlated when responding to the personality measure becomes a complex task, such as when taking a personality test in an applicant setting when warnings about faking are present or a forced-choice format is used. We detected no significant correlations between cognitive ability and four of the Big Five personality in the present study, suggesting that the speeding manipulation did not render the personality measure more cognitively demanding. The exception was Openness, which  2010 Hogrefe Publishing

was significantly correlated with cognitive ability in all conditions except for honest-speeded. This is likely due to overlap between the two constructs of cognitive ability and Openness – the latter sometimes referred to as Intellectance (Costa & McCrea, 1992). On the other hand, there was a significant relationship between cognitive ability and IM in the faked-speeded condition. The IM measure may have more complex items (e.g., double-barreled or job-irrelevant items) that were difficult to fake under time pressure. Clearly, future tests of resource allocation theory (Ackerman, 1986) should take participant cognitive ability into account, as well as the complexity of the task.

Practical Implications The most startling implication from the current study is what may happen to individuals of high g under speeded conditions in selection situations when IM scales are used in the decision-making process. According to our results, when faking is expected and completion time is restricted, if high IM scores are used to remove individuals from the selection pool, a higher proportion of high g compared to lower g individuals will be removed. Given the high validity of g in predicting job performance across a range of jobs (Schmidt & Hunter, 1998), this is surely not an advantageous situation as it will likely result in lower overall utility of the selection system. However, it is important to note that the boundary conditions of the speeding phenomenon have not properly been delineated (see Future Research section).

Limitations This study has several limitations. One limitation was that we used a student sample to test hypotheses that were designed to generalize to an applicant selection context. We believe that the amount of experimental control that Journal of Personnel Psychology 2010; Vol. 9(3):126–137

134

S. Komar et al.: Speeding Personality Measures and Faking

we gained by using students (i.e., increased internal validity) offsets possible issues with external validity issues in that regard. Moreover, manipulation checks suggested that students were sufficiently motivated to create a positive impression. Another limitation that could be addressed in future research is to examine the effect of speed in the context of faking on criterion-related validity. Although our result, showing no main effect for speed, suggests a possible null effect on criterion-related validity, it does not presuppose such a null effect.

Future Research Boundary conditions for the speed effect in the context of faking should be examined. One such boundary condition is the presence of warnings that faking can be identified and/or penalized (Dwight & Donovan, 2003). Warnings may increase the cognitive load associated with faking on a noncognitive questionnaire if participants are trying to elevate their scores without being detected (Vasilopoulos et al., 2005). Accordingly, warnings should reduce faking, particularly under a speeded administration. The other such boundary condition concerns the format of the test. The increased complexity of forced-choice, situational judgment, or subtle items on conditional reasoning measures may increase cognitive load, making it more likely that speeding will result in reduced faking. Future research should include the intervening variables noted in Figure 1. In this way, both the direct and indirect effects of test speeding can be examined in relation to outcome variables such as test motivation, self-doubts regarding test-taking, and dislike of tests. Furthermore, future research should examine how these reactions relate to behavioral outcomes, such as accepting job offers and withdrawing from the applicant pool.

Conclusion The current study is the first to find evidence of a possible influence of speed on IM and g in a faking context. When faking is expected and IM scores are used to remove individuals from the selection pool, speeded conditions of test administration will result in high cognitive ability individuals being removed from the applicant pool. Therefore, manipulating time to complete personality measures is not suggested at the present time.

Acknowledgment This research was supported by a grant from the Social Sciences and Humanities Research Council of Canada to Chet Robie and Simon Taggar.

Journal of Personnel Psychology 2010; Vol. 9(3):126–137

References Ackerman, P. L. (1986). Individual differences in information processing: An investigation of intellectual abilities and task performance during practice. Intelligence, 10, 101–139. Aitkens, H. A., Thorndike, E. L., & Hubbell, E. (1902). Correlations amongst perceptive and associative processes. Psychological Review, 9, 374–382. Ambrose, M. L., & Rosse, J. G. (2003). Procedural justice and personality testing. Group and Organization Management, 28, 502–526. Arthur, W., Jr., Glaze, R. M., Villado, A. J., & Taylor, J. E. (2009). Unproctored Internet-based tests of cognitive ability and personality: Magnitude of cheating and response distortion. Industrial and Organizational Psychology, 2, 39–45. Arvey, R. D., Strickland, W., Drauden, G., & Martin, C. (1990). Motivational components of test taking. Personnel Psychology, 43, 695–716. Bargh, J. (1989). Conditional automaticity: Varieties of automatic influence in social perception and cognition. In J. S. Uleman & J. A. Bargh (Eds.), Unintended thought (pp. 3–51). New York, NY: Guilford Press. Barrick, M. R., Mount, M. K., & Judge, T. A. (2001). Personality and performance at the beginning of the new millennium: What do we know and where do we go next? International Journal of Selection and Assessment, 9, 9–30. Bible, J. D. (1990). When employers look for things other than drugs: The legality of AIDS, genetic, intelligence, and honesty testing in the workplace. Labor Law Journal, 41, 195–213. Brown, R. D., & Cothern, C. M. (2002). Individual differences in faking integrity tests. Psychological Reports, 91, 691–702. Burns, G. N., & Christiansen, N. D. (2006). Sensitive or senseless: On the use of social desirability measures in selection and assessment. In R. L. Griffith & M. H. Peterson (Eds.), A closer examination of applicant faking behaviour (pp. 113–148). Greenwich, CT: Information Age Publishing. Costa, P. T., & McCrae, R. R. (1992). Revised NEO Personality Inventory (NEO–PI–R) and the NEO Five-Factor Inventory (NEO-FFI). Odessa, FL: Psychological Assessment Resources. Crant, J. M., & Bateman, T. S. (1990). An experimental test of the impact of drug-testing programs on potential job applicants’ attitudes and intentions. Journal of Applied Psychology, 75, 127–131. Cronbach, L. J., & Warrington, W. G. (1951). Time limit tests: Estimating their reliability and degree of speeding. Psychometrika, 16, 167–188. Dawson, J. F., & Richter, A. W. (2006). Probing three-way interactions in moderated multiple regression: Development and application of a slope difference test. Journal of Applied Psychology, 91, 917–926. DeShon, R. P., Brown, K. G., & Greenis, J. L. (1996). Does selfregulation require cognitive resources? Evaluation of resource allocation models of goal setting. Journal of Applied Psychology, 81, 595–608. Donlon, T. F. (1980). An annotated bibliography of studies of test speededness, GRE Board Research Report GREB No. 76-9R. Princeton, NJ: Educational Testing Service. Dwight, S. A., & Donovan, J. J. (2003). Do warnings not to fake reduce faking? Human Performance, 16, 1–23. Ellingson, J. E., Smith, D. B., & Sackett, P. R. (2001). Investigating the influence of social desirability on personality factor structure. Journal of Applied Psychology, 86, 122–133. Evans, F. R., & Reilly, R. R. (1972). A study of speededness as a source of test bias. Journal of Educational Measurement, 9, 123–131.

 2010 Hogrefe Publishing

S. Komar et al.: Speeding Personality Measures and Faking

Freeman, F. S. (1928). Power and speed: Their influence upon intelligence test scores. Journal of Applied Psychology, 12, 631–637. Goffman, E. (1959). The presentation of self in everyday life. New York, NY: Doubleday. Goldberg, L. R., Johnson, J. A., Eber, H. W., Hogan, R., Ashton, M. C., Cloninger, C. R., & Gough, H. C. (2006). The International Personality Item Pool and the future of public-domain personality measures. Journal of Research in Personality, 40, 84–96. Hausknecht, J. P., Day, D. V., & Thomas, S. C. (2004). Applicant reactions to selection procedures: An updated model and meta-analysis. Personnel Psychology, 57, 639–683. Holden, R. R. (1995). Response latency detection of fakers on personnel tests. Canadian Journal of Behavioural Science, 27, 343–355. Holden, R. R. (1998). Detecting fakers on a personnel test: Response latencies versus a standard validity scale. Journal of Social Behavior and Personality, 13, 387–398. Holden, R. R., & Hibbs, N. (1995). Incremental validity of response latencies for detecting fakers on a personality test. Journal of Research in Personality, 29, 362–372. Hooper, A. C., Cullen, M. J., & Sackett, P. R. (2006). Operational threats to the use of SJTs: Faking, coaching, and retesting issues. In J. A. Weekley & R. E. Ployhart (Eds.), Situational judgment tests: Theory, measurement, and application (pp. 205–232). Mahwah, NJ: Erlbaum. Hsu, C. (2004). The testing of America. U.S. News and World Reports, 137, 68–69. Jackson, D. N., Wroblewski, V. R., & Ashton, M. C. (2000). The impact of faking on employment tests: Does forced choice offer a solution? Human Performance, 13, 371–388. John, O. P., & Srivastava, S. (1999). The Big-Five trait taxonomy: History, measurement, and theoretical perspectives. In L. A. Pervin & O. P. John (Eds.), Handbook of personality: Theory and research (2nd ed., pp. 102–138). New York, NY: Guilford Press. Kanfer, R., & Ackerman, P. L. (1989). Motivation and cognitive abilities: An integrative/aptitude-treatment interaction approach to skill acquisition. Journal of Applied Psychology, 74, 657–690. Komar, S., Brown, D. J., Komar, J. A., & Robie, C. (2008). Faking and the validity of conscientiousness: A Monte Carlo investigation. Journal of Applied Psychology, 93, 140–154. Kuncel, N. R., & Tellegen, A. (2009). A conceptual and empirical reexamination of the measurement of the social desirability of items: Implications for detecting desirable response style and scale development. Personnel Psychology, 62, 201–228. Lanyon, R. I., & Carle, A. C. (2007). Internal and external validity of scores on the Balanced Inventory of Desirable Responding and Paulhus Deception Scales. Educational and Psychological Measurement, 67, 859–876. LeBreton, J. M., Barksdale, C. D., Robin, J., & James, L. R. (2007). Measurement issues associated with conditional reasoning tests: Indirect measurement and faking issues. Journal of Applied Psychology, 92, 1–16. Lord, R. G., & Levy, P. E. (1994). Moving from cognition to action: A control theory perspective. Applied Psychology: An International Review, 43, 335–398. McCarthy, J. M., & Goffin, R. D. (2003). Is the Test Attitude Survey psychometrically sound? Educational and Psychological Measurement, 63, 446–464. McHenry, J. J., Hough, L. M., Toquam, J. L., Hanson, M. A., & Ashworth, S. A. (1990). Project A validity results: The relationship between predictor and criterion domains. Personnel Psychology, 43, 335–354.  2010 Hogrefe Publishing

135

Murphy, K. R., & Dzieweczynski, J. L. (2005). Why don’t measures of broad dimensions of personality perform better as predictors of job performance? Human Performance, 18, 343–357. Ones, D. S., Viswesvaran, C., & Reiss, A. D. (1996). Role of social desirability and faking on personality testing for personnel selection: The red herring. Journal of Applied Psychology, 81, 660–679. Paulhus, D. L. (1991). Measurement and control of response bias. In J. P. Robinson, P. R. Shaver & L. S. Wrightsman (Eds.), Measures of personality and social psychological attitudes (pp. 17–59). New York, NY: Academic Press. Paulhus, D. L. (2001). Socially desirable responding: The evolution of a construct. In H. Braun, D. N. Jackson & D. E. Wiley (Eds.), The role of constructs in psychological and educational measurement (pp. 67–88). Mahwah, NJ: Erlbaum. Pauls, C. A., & Crost, N. W. (2005). Cognitive ability and selfreported efficacy of self-presentation predict faking on personality measures. Journal of Individual Differences, 26, 194–206. Payne, J. W, Bettman, J. R., & Johnson, E. J. (1993). The adaptive decision maker. New York, NY: Cambridge University Press. Ployhart, R. E., & Ryan, A. M. (1997). Toward an explanation of applicant reactions: An examination of organizational justice and attribution frameworks. Organizational Behaviour and Human Decision Processes, 72, 308–335. Robie, C., Brown, D. J., & Beaty, J. C. (2007). Do people fake on personality inventories? A verbal protocol analysis. Journal of Business and Psychology, 21, 489–509. Robie, C., Curtin, P. J., Foster, T. C., Phillips, H. L. IV, Zbylut, M., & Tetrick, L. E. (2000). The effect of coaching on the utility of response latencies in detecting fakers on a personality measure. Canadian Journal of Behavioural Science, 32, 226–233. Robie, C., Tuzinski, K. A., & Bly, P. R. (2006). A survey of assessor beliefs and practices related to faking. Journal of Managerial Psychology, 21, 669–681. Robie, C., Zickar, M. J., & Schmit, M. J. (2001). Measurement equivalence between applicant and incumbent groups: An IRT analysis of personality scales. Human Performance, 14, 187–207. Rosse, J. G., Stecher, M. D., Miller, J. L., & Levin, R. A. (1998). The impact of response distortion on preemployment personality testing and hiring decisions. Journal of Applied Psychology, 83, 634–644. Ryan, A. M., McFarland, L., Baron, H., & Page, R. (1999). An international look at selection practices: Nation and culture as explanations for variability in practice. Personnel Psychology, 52, 359–391. Rynes, S., Bretz, R., & Gerhart, B. (1991). The importance of recruitment in job choice: A different way of looking. Personnel Psychology, 44, 487–521. Sackett, P. R., & Lievens, F. (2008). Personnel selection. Annual Review of Psychology, 59, 415–450. Schmidt, F. L., & Hunter, J. E. (1998). The validity and utility of selection methods in personnel psychology: Practical and theoretical implications of 85 years of research findings. Psychological Bulletin, 124, 262–274. Schmitt, N., & Oswald, F. L. (2006). The impact of corrections for faking on the validity of noncognitive measures in selection settings. Journal of Applied Psychology, 91, 613–621. Sweller, J., Van Merrie¨nboer, J. J. G., & Paas, F. G. W. C. (1998). Cognitive architecture and instructional design. Educational Psychology Review, 10, 251–296. Tull, K. T. (1998). The effects of faking behaviour on the prediction of sales performance using the Guilford Zimmerman Temperament Survey and the NEO Five Factor Journal of Personnel Psychology 2010; Vol. 9(3):126–137

136

S. Komar et al.: Speeding Personality Measures and Faking

Inventory, Unpublished doctoral dissertation, University of Akron, Akron, OH. Vasilopoulos, N. L., Cucina, J. M., Dyomina, N. V., Morewitz, C. L., & Reilly, R. R. (2006). Forced-choice personality tests: A measure of personality and cognitive ability? Human Performance, 19, 175–199. Vasilopoulos, N. L., Cucina, J. M., & McElreath, J. M. (2005). Do warnings of response verification moderate the relationship between personality and cognitive ability? Journal of Applied Psychology, 90, 306–322. Vasilopoulos, N. L., Reilly, R. R., & Leaman, J. A. (2000). The influence of job familiarity and impression management on self-report measure scale scores and response latencies. Journal of Applied Psychology, 85, 50–64. Wiechmann, D., & Ryan, A. M. (2003). Reactions to computerized testing in selection contexts. International Journal of Selection and Assessment, 11, 215–229. Wonderlic, E. F. (1992). Wonderlic Personnel Test user’s manual. Libertyville, IL: E. F. Wonderlic.

Journal of Personnel Psychology 2010; Vol. 9(3):126–137

Zickar, M. J., Gibby, R. E., & Robie, C. (2004). Uncovering faking samples in applicant, incumbent, and experimental data sets: An application of mixed model item response theory. Organizational Research Methods, 7, 168–190. Zickar, M. J., & Robie, C. (1999). Modeling faking good on personality items: An item-level analysis. Journal of Applied Psychology, 84, 551–563.

Shawn Komar Department of Psychology 200 University Avenue West Waterloo Ontario, N2L 3G1 Canada E-mail [email protected]

 2010 Hogrefe Publishing

S. Komar et al.: Speeding Personality Measures and Faking

Appendix A The following were the instructions given to participants assigned to ‘‘honest’’ experimental cells: Before I hand out the personality test I want to remind you that the test you are about to take is one that is frequently used by employers to select employees for all types of jobs (e.g., managerial jobs, salespersons, police officers, etc.). Please answer the following questions on the personality test as honestly as possible. Your answers will remain completely anonymous. The nature of the project requires that you answer the following questions as honestly as possible, so please provide as accurate answers as you can. The following were the instructions given to participants assigned to ‘‘fake’’ experimental cells: Before I hand out the personality test I want to remind you that the test you are about to take is one that is frequently used by employers to select employees for all types of jobs (e.g., managerial jobs, salespersons, police officers, etc.). When answering the questions on the personality test imagine that you are applying for the following job. Please take your time to read this job advertisement. Has everyone completed reading the job advertisement? Again, answer the questions on the personality test as if you were applying for the job you just read about. Present yourself in a way that will get you hired. To make this situation more like an application situation, we are offering an incentive. Those of you who score in the top 15% on this test will receive $20 in addition to the credit for your participation in this experiment. Keep in mind that your answers will be kept completely anonymous. The following were the instructions given to participants assigned to ‘‘unspeeded’’ experimental conditions: There is no time limit on this test. Note that the test has a front and a back side. Please do not skip questions (i.e., finish all the questions on the first side of the page before continuing on to the following page). Please write your answers on both the test form and the scantron form. You may begin.

137

write your answers on both the test form and the scantron form. You may begin.

Appendix B Job advertisement Advertisement number: 2354442 Title: Business general manager (Business development manager) (NOC: 0014) Terms of employment: Permanent, full time, day. Salary: $70,000.00–$100,000.00 Anticipated start date: As soon as possible. Location: Toronto, Ontario (one vacancy). Skill requirements: Education: Completion of university. Languages: Speak English, read English, write English. Essential skills: Reading text, document use, numeracy, writing, oral communication, working with others, problem solving, decision making, critical thinking, job task planning and organizing, significant use of memory, finding information, computer use, continuous learning. Other information: The successful candidate will have a BBA degree and a passion for a ‘‘green friendly business.’’ Solid understanding of how to motivate staff, teamwork, and strategy are beneficial. Comfortable in a fast-paced, innovative industry. Employer: Clean Environment Inc. – a leading R and D firm specializing in the creation and application of industrial air filtration systems (2005 revenues of greater than $15 million and growing year over year). How to Apply: Please apply for this job only in the manner specified by the employer. Failure to do so may result in your application not being properly considered for the position. By E-mail: [email protected] Business profile: Technology firm in the Toronto area. Advertised until: 2007/5/15

The following were the instructions given to participants assigned to ‘‘speeded’’ experimental conditions: You will be given 10 min to complete the test. Note that the test has a front and a back side. Please do not skip questions (i.e., finish all the questions on the first side of the page before continuing on to the following page). I will notify you as to how much time you have remaining in 2-min intervals. Please  2010 Hogrefe Publishing

Journal of Personnel Psychology 2010; Vol. 9(3):126–137