The Specificity of Maternal Parenting Behavior and Child ... - CiteSeerX

Journal of Family Psychology 2008, Vol. 22, No. 2, 181–192

Copyright 2008 by the American Psychological Association 0893-3200/08/$12.00 DOI: 10.1037/0893-3200.22.2.181

The Specificity of Maternal Parenting Behavior and Child Adjustment Difficulties: A Study of Inner-City African American Families Deborah J. Jones

Rex Forehand, Aaron Rakow, Christina J. M. Colletti, and Laura McKee

University of North Carolina at Chapel Hill

University of Vermont

Alecia Zalot University of North Carolina at Chapel Hill The specificity of the association between 2 parenting behaviors (warmth and supervision) and 2 indicators, aggressive behavior and depressive symptoms, of major child outcomes (externalizing problems and internalizing problems) was examined among 196 inner-city African American mothers and their school age children. Given the growing number of African American families affected by HIV/AIDS and demonstrated compromises in parenting associated with maternal infection, the moderating role of maternal HIV/AIDS was also examined. Findings from longitudinal analyses supported the specificity of maternal warmth but not of maternal supervision. Maternal warmth was a stronger predictor of decreases in child aggressive behavior than of decreases in depressive symptoms. In addition, maternal warmth was a stronger predictor of decreases in aggressive behavior than was maternal supervision. Parenting specificity was not moderated by maternal HIV/AIDS. Clinical implications and future research directions are discussed. Keywords: parenting specificity, maternal HIV/AIDS, child adjustment, African American

Consistent with the importance of studying child development within context, the family has been a primary focus of studies in the area of developmental psychopathology (see Cummings, Davies, & Campbell, 2000, for a review). Within this literature, authoritative parenting behavior, or parenting behavior characterized by a balance of warmth (i.e., support, involvement, open communication) and supervision (i.e., awareness of child’s activities inside and outside the home, e.g., television programs watched, homework completed, school activities, peer activities), has been identified as the ideal context to promote child adjustment (Baumrind, 1991). Although a robust literature now docu-

ments the link between both parental warmth and supervision and a broad range of child outcomes (e.g., Davidow & Grusec, 2006), relatively little attention has been devoted to the study of the specificity of these associations (Cummings et al., 2000). That is, we do not know the extent to which a specific parenting behavior (e.g., supervision) has a unique (i.e., direct) effect on a specific child outcome (e.g., aggressive behavior), after controlling for other parenting and child behaviors (Caron, Weiss, Harris, & Catron, 2006; also see O’Connor, 2002, for a review). Similarly, we know little about the extent to which one parenting behavior (e.g., supervision) has differential (i.e., stronger) effects on two or more child outcomes (e.g., aggressive behavior and depressive symptoms; Caron et al., 2006; O’Connor, 2002). Notably, well-established theories of parent– child interactions, such as the coercion hypothesis, do posit causal links between specific parenting behaviors (i.e., parental failure to follow through with commands and requests in response to noncompliance) and specific child outcomes (i.e., future noncompliance and, ultimately, oppositional and conduct disordered behavior; see Granic & Patterson, 2006, for a review). Although a long history of empirically supported parenting programs has been built on such hypotheses (see McMahon, Wells, & Kotler, 2006, for a review), a review of the empirical literature suggests that the evidence for parenting specificity may be far less clear (O’Connor, 2002). Consistent with the concept of multifinality (Cummings et al., 2000), specific parenting behaviors, such as parental warmth, have been linked with mul-

Deborah J. Jones and Alecia Zalot, Department of Psychology, University of North Carolina at Chapel Hill; Rex Forehand, Aaron Rakow, Christina J. M. Colletti, and Laura McKee, Department of Psychology, University of Vermont. Funding for this project was provided in part by the Centers for Disease Control and Prevention and the William T. Grant Foundation. Additional support for data analyses and manuscript preparation were provided by the Institute for Behavioral Research at the University of Georgia and a Research and Study Leave awarded to Deborah J. Jones from the University of North Carolina at Chapel Hill. Appreciation is expressed to Edward Morse and Patricia Simon Morse for their role in data collection and J. J. Bau for his assistance with data analyses. Correspondence regarding this article should be addressed to Deborah J. Jones, Department of Psychology, Davie Hall, CB#3270, Chapel Hill, North Carolina 27599-3270. E-mail: [email protected]

181

182

JONES ET AL.

tiple child outcomes in the literature, including indicators of externalizing (e.g., Lee & Gotlib, 1991; Masten & Coatsworth, 1998; Shaw et al., 1998) and internalizing (e.g., Garber, Robinson, & Valentiner, 1997; Hammen, Shih, & Brennan, 2004) problems (i.e., non-specific effects). Moreover, single child outcomes, such as indicators of externalizing problems, have been linked to multiple parenting behaviors, including both warmth (e.g., Lee & Gotlib, 1991; Masten & Coatsworth, 1998; Shaw et al., 1998) and supervision (e.g., Chilcoat, Breslau, & Anthony, 1996; Dishion, Patterson, Stoolmiller, & Skinner, 1991; Gray & Steinberg, 1999), a concept referred to as equifinality (Cummings et al., 2000). Although these findings seem to contradict the theoretical underpinnings of parent training programs by suggesting relatively non-specific effects of parenting, methodological and analytical variation across studies precludes such a definitive conclusion. Importantly, direct tests of parenting specificity must first measure at least two parenting behaviors and at least two child outcomes, then examine the following (Caron et al., 2006; O’Connor, 2002): (a) whether one of the parenting behaviors is associated with one of the child outcomes, after controlling for co-occurring parenting behaviors and the other child outcome (i.e., unique association); and (b) whether the magnitude or direction of the association between one of the parenting behaviors and one of the child outcomes differs significantly from the association between the same parenting behavior and another child outcome (i.e., differential association). To date, only one study directly tested parenting specificity with the aforementioned criteria (Caron et al., 2006). In their cross-sectional study of both Caucasian and African American families, Caron and colleagues (2006) examined the interrelationship of three dimensions of parenting behavior (warmth, behavioral control, and psychological control), and two domains of child adjustment difficulties (internalizing and externalizing problems). Findings revealed that parental behavioral control (i.e., stating a consequence, suggesting alternatives, explaining why a rule is enforced, limit setting, and monitoring), but not parental warmth (i.e., positive comments about the child or the child’s behavior, affirmations, active speaking during the task) or psychological control (i.e., guilt induction, hostile tone, emotional overinvolvement, and manipulative threats), was uniquely and differentially associated with measures of internalizing and externalizing problems. That is, behavioral control was uniquely associated with internalizing problems (after controlling for the other parenting behaviors and externalizing problems) and with externalizing problems (after controlling for the other parenting behaviors and internalizing problems). In addition, behavioral control was differentially associated with the two child outcomes, such that greater behavioral control was associated with higher levels of internalizing problems but with lower levels of externalizing problems. The findings of Caron and colleagues (2006) represented an important first step in the study of parenting specificity; however, several methodological issues limit the conclusions that can be drawn from their research. First, they utilized a cross-sectional design, precluding any determina-

tion about the direction of associations. In addition, although Caron et al. included both Caucasian and African American youths in their study, they did not examine the role of race/ethnicity. Given that the need for prevention and intervention programs with inner-city ethnic minority children is being increasingly recognized (e.g., Brody et al., 2006; Forehand, Miller, Armistead, Kotchick, & Long, 2004; Reid, Webster-Stratton, & Beauchaine, 2001), an examination of the specificity of parenting behavior within an inner-city African American sample seems particularly relevant. The current longitudinal study aimed to replicate and extend the study by Caron and colleagues (2006) by examining the specificity of two dimensions of maternal parenting behavior (warmth and supervision) and a child outcome measure of externalizing problems (i.e., aggressive behavior) and internalizing problems (i.e., depressive symptoms) among inner-city African American families with a school age child. In regard to unique specificity, we hypothesized that warmth would be more strongly associated with changes over time in both aggressive behavior and depressive symptoms than would supervision. It is important to note that our hypothesis is not consistent with the findings of Caron et al.; however, substantial evidence in the literature has suggested that parental warmth may be the most important parenting variable for protecting youths residing in the difficult conditions (e.g., inner-city environments, poverty) that characterized our sample (Kim-Cohen, Moffitt, & Caspi, 2004). For example, in their review of supervision and monitoring, Dishion and McMahon (1998) stressed that the foundation for parenting skills, such as supervision and effective behavior change, is the parent– child relationship (i.e., warmth). Similarly, after a review of the literature, Masten and Coatsworth (1998) concluded that parental warmth is the critical variable for promoting resilience in children living in challenging environments. Second, we predicted that both parental warmth and parental supervision would have a differential association with aggressive behavior and depressive symptoms. Specifically, we expected both variables to contribute more to changes (i.e., decreases or smaller increases) over time in aggressive behavior than in depressive symptoms. The association of parenting variables and child aggressive behavior has received more attention than the association of parenting variables and either child depressive symptoms or anxious symptoms (for a review of the literature on specificity, see McKee, Colletti, Rakow, Jones, & Forehand, 2007; for a review of the treatment literature, see Barmish & Kendall, 2005; Horowitz & Garber, 2006; McMahon, Wells, & Kotler, 2006). Specifically, behavioral parenting programs for aggressive and other acting out behaviors typically include a warmth component and, occasionally, a supervision component. It is through this combination of parenting skills that parents learn to structure the child’s behavior (McMahon et al., 2006). Far fewer empirical studies have examined the role of parenting interventions in the treatment of children’s internalizing problems (see Diamond & Josephson, 2005; Horowitz & Garber, 2006, for reviews). Although some

SPECIFICITY OF MATERNAL PARENTING BEHAVIOR

evidence has suggested that the case for the inclusion of parents in interventions for internalizing disorders may be building, the evidence is mixed and the studies with outcome data are too few in number to make any definitive conclusions (Diamond & Josephson, 2005). Consistent with the contextual models guiding the study of developmental psychopathology (Cummings et al., 2000), our investigation of parenting specificity in urban African American families considered an additional contextual factor, maternal HIV/AIDS. African American youths are more likely than children in any other group to have a mother with HIV/AIDS (Anderson & Smith, 2003; Centers for Disease Control and Prevention [CDC], 2004a; 2004b), and HIV/AIDS has been associated with compromises in maternal parenting behavior (e.g., Antle, Wells, & Goldie, 2001; Kotchick et al., 1997; Tompkins, Henker, & Whalen, 1999), which, in turn, have been associated with greater problem behaviors for their non-infected children (e.g., Forehand et al., 1998). These findings suggest that the unique and differential associations of parental warmth and supervision with child aggressive behavior and depressive symptoms may be attenuated in families in which mothers are infected with HIV/AIDS. Accordingly, we hypothesized an interaction between each parenting variable and maternal HIV status.

183

Method Overview The current study represents secondary analyses of data from the Family Health Project, a longitudinal investigation of the psychosocial functioning of low-income, inner-city African American mothers, over one-third of whom were infected with HIV/AIDS, and their non-infected youths (Family Health Project Research Group, 1998). At each annual assessment, mother– child dyads participated in two separate interviews: the sociological interview, which assessed demographic information, and the psychological interview, which assessed major study variables. The current analyses focus on the first and second assessments of the Family Health Project.

Participants Participants for the current study were 196 African American mothers (71 HIV-infected, 125 non-infected) and their 6- to 11-year-old non-infected children who completed two assessments separated by 12 months. An additional 51 participants completed only the first assessment and are not included in the current analyses. The majority of participants lived in government housing projects characterized by overcrowding, high levels of poverty, and crime (see Table 1, which presents information on

Table 1 Sample Demographics at Time 1 (n ⫽ 196) for the Overall Sample and the HIV Non-Infected and HIV Infected Subsamples Overall sample Variable Child Age (years) Gender % girls % boys Mother Age (years) Education Less than high school High school or GED Some vocational school Some college Employment statusa Marital statusb Family Monthly Income Childrenc Adultsd 0 1 2 3 4 or more a

M

SD

8.86

1.69

HIV non-infected %

M

SD

8.89

1.63

54 46 32.50

5.58

33.33

$471 1.51

SD

8.81

1.70

5.60

31.25

$502 1.53

5.33 34 41 8 17 42 11

$640 2.67 62 17 14 2 5

%

55 45

46 30 10 14 47 22 $717 3.27

64 18 12 2 4

M

53 47

42 34 9 15 45 18 $689 3.05

HIV infected %

$410 1.40 68 21 7 1 3

Percent employed. Percent married. Number of children of mother living in household. d Percent of households with adults other than mother: 0 ⫽ no other adults; 1 ⫽ 1 other adult; 2 ⫽ 2 other adults; 3 ⫽ 3 other adults; 4 or more ⫽ 4 or more adults. b c

184

JONES ET AL.

the child, mother, and family for the complete sample, the non-infected sample, and the HIV-infected sample). On average, mothers were 32.5 years of age, nearly half (42%) failed to complete high school, and the minority (18%) were married. Children averaged 8.86 years of age at the first assessment and approximately half (54%) were girls. Families in the study reported an average monthly income of $689. Specific eligibility criteria for the Family Health Project were as follows: Mothers, who ranged in age from 18 to 45 years, had to report no intravenous drug use for at least 6 months prior to the first assessment. In addition, they had to have at least one biological child in the 6- to 11year-old age range. Participants in the maternal HIV/AIDS sample were recruited through the primary public health clinic in New Orleans (93%), as well as from private physician’s offices (7%), over a 2-year period. During well visits, a project staff member approached mothers who appeared to meet study inclusion requirements, and then he or she explained the study, confirmed eligibility, and scheduled a baseline assessment session. In order to study mothers who were in a relatively advanced stage of illness, we limited the HIV/ AIDS sample to those mothers with a CD4 count under 600. A stratified random comparison sample was drawn from the same zip code areas in which the maternal HIV/AIDS sample resided. All of these zip codes were from inner-city New Orleans. Zip codes of the HIV-infected sample were continuously tracked, and public schools were identified that served these areas. The comparison sample was recruited through five of the six public schools serving the targeted zip code areas. Recruitment occurred in two waves spanning 2 school years, with an equal number of participants recruited each year. In each of the two waves, letters describing the study and inviting families to participate were sent to the homes of 30 African American mothers at each school. School personnel were asked to randomly select families to receive letters. In each wave, the first 15 mothers at each school to return a reply card constituted the sample. Mothers in this group identified themselves as noninfected in the first interview. In the HIV/AIDS sample, 95% of the mother– child dyads who were approached agreed to, and did, participate in each of two scheduled interviews. For the non-infected sample, 94% of the mother– child dyads returned the reply card indicating an interest in participating. Of the mother– child dyads selected for participation on the basis of the order of the return of their cards, 100% participated in an initial sociological interview, and 99% of these participated in the subsequent psychological interview.

Measures Selection of culturally sensitive and appropriate instruments consisted of several steps. These included selecting constructs, discussing the constructs within focus groups, selecting and pilot testing measures with a group of African American mothers, and verifying the appropriateness of the major constructs to be assessed with two more focus groups of African American mothers living in New Orleans. For

more information, see Family Health Project Research Group (1998). Demographic information and maternal HIV/AIDS diagnosis. The following demographic information was obtained from the mothers and children: age of child and mother; gender of the child; education of mother (1 ⫽ less than 9th grade; 2 ⫽ less than 12th grade; 3 ⫽ high school degree; 4 ⫽ some college or college degree); employment status and marital status of mother; average family income from all sources received monthly; number of children in the household; and percentage of families with 0, 1, 2, 3, and 4 or more adults other than the mother living in the family. For those in the HIV/AIDS group, a diagnosis of maternal HIV/AIDS was confirmed via medical chart review and Centers for Disease Control and Prevention (CDC, 1992) staging procedures, the standard at the time of data collection, yielding 54%, 21%, and 25% classified as asymptomatic, symptomatic, and AIDSdiagnosed, respectively. Average time since diagnosis was 3.1 years (SD ⫽ 2.1 years). Maternal warmth. Mother report on the short form of the Interaction Behavior Questionnaire (IBQ; Prinz, Foster, Kent, & O’Leary, 1979) was used to assess warmth and support in the mother– child relationship. This form consists of the 20 items that have the highest phi coefficients and the highest item-to-total correlations with the 75 items in the original IBQ. The short form correlates .96 with the longer version. The IBQ has a long history as a valid measure of parental warmth and support. Parental warmth should be (a) more evident in parent–youth relationships that are characterized by positive, rather than negative, interactions; (b) associated with youth psychosocial adjustment; and (c) less evident when parents experience psychological or physical illness (see Armistead, Klein, & Forehand, 1995; Masten & Coatsworth, 1998). The IBQ does discriminate distressed from nondistressed parent–youth dyads (Prinz et al., 1979), and it predicts short- and long-term (i.e., 6 years) adjustment from adolescence to young adulthood (e.g., Summers, Forehand, Armistead, & Tannenbaum, 1998). Furthermore, and of importance, our research with school age African American children provides support for the measure: Higher IBQ scores (more warmth) are associated with more resilience among children living in distressed families (Dutra, Forehand, & Armistead, 2000); physically ill mothers report lower IBQ scores than do non-ill mothers (Kotchick et al., 1997); and higher IBQ scores are inversely related to indicators of internalizing and externalizing symptoms of school age children in single-parent families (Brody, Dorsey, Forehand, & Armistead, 2002). Sample items, which may be endorsed as True or False, include “You enjoy spending time with your child” and “You think you and your child get along well with each other.” Prinz and colleagues (1979) reported adequate internal consistency and discriminant validity. A confirmatory factor analysis indicated that 19 of the 20 items loaded on a single construct; accordingly, only these 19 items were included in the measure for data analysis. The alpha coefficient for these 19 items was .83. Scores can range from 0 to 19, with higher scores, after reverse scoring, indicating


greater warmth and support in the mother– child relationship. Maternal supervision. The Monitoring and Control Questionnaire (MCQ) was used to assess the extent to which a mother supervised her child’s behavior. This is a particularly important parenting skill for school age children who live in high-risk (e.g., inner-city) environments (Brody et al., 2001; Dishion & McMahon, 1998; Gorman-Smith & Tolan, 1998). This 17-item scale is based on monitoring measures used by Patterson and Stouthamer-Loeber (1984) and by Steinberg, Lamborn, Dornbusch, and Darling (1992). It assesses parents’ perceptions of their knowledge about various aspects of their children’s lives, including school performance, television and movies watched, and peer group activities. Our research, as well as that of others, provides evidence that the MCQ is a valid measure of the supervision construct with school age African American children. Monitoring should (a) occur more in high-risk than low-risk environments; (b) be more protective for youths in high-risk, rather than low-risk, environments; (c) occur less often when parents experience psychological or physical illness; and (d) be associated with lower levels of problem behaviors and associated difficulties (see Dishion & McMahon, 1998). Mothers living in high-risk (i.e., inner-city) environments do report higher scores on the MCQ than those of mothers living in low-risk (i.e., rural) environments (Armistead, Forehand, Brody, & Maguen, 2002); the MCQ is associated with fewer child problem behaviors in high-risk, but not low-risk, environments (Armistead et al., 2002); mothers who are physically ill (i.e., HIV-infected) report lower MCQ scores than do non-infected mothers (Kotchick et al., 1997), and children of the former group of mothers have more problem behaviors than do those of the latter group of mothers (Forehand et al., 1998); higher scores on the MCQ are part of a parenting construct that protects children from affiliation with deviant peers in disadvantaged neighborhoods (Brody et al., 2001). Items are rated on a 4-point Likert scale ranging from 1 (never) to 4 (always). Sample items include “How often do you know about where [target child] is and what s/he is doing when away from home?” “How often do you know about television programs that [target child] watches?” and “How often do you know about what his/her grades are?” Scores can range from 17 to 68, with higher scores indicating higher levels of maternal monitoring. For the present sample, a factor analysis indicated that all 17 items loaded. The alpha coefficient for this scale was .86. Youth externalizing problems. Child-reported aggressive behavior, assessed by the Aggression subscale from the Youth Self-Report form (Achenbach, 1991) of the Child Behavior Checklist (Achenbach, 1991), served as an indicator of externalizing problems. This subscale was selected because it assesses the types of behaviors typically displayed by children in the age range included in this study, and it has acceptable reliability and validity data (Achenbach, 1991). However, it has not been standardized with children as young as some of those included in this investigation (i.e., those under age 11 years); thus, we initially

185

conducted a factor analysis on the scale. All 19 items of the Aggression subscale loaded at .40 or greater and were retained. The alpha coefficient for this subscale with the current sample was .86. Scores on the scale can range from 0 to 47. To provide additional evidence for use of the Aggression subscale of the Youth Self-Report measure with children under age 11, we conducted two sets of analyses. First, we computed an alpha coefficient and mean score for aggressive behavior for younger (under age 11.0) and older (11.0 and older) children. The alpha coefficients were identical at Time 1 (.85) and very similar at Time 2 (younger ⫽ .84 and older ⫽ .86). The means did not differ at Time 1 (younger ⫽ 6.04 and older ⫽ 7.0), t(194) ⫽ .96; or at Time 2 (younger ⫽ 5.85 and older ⫽ 5.49) t(194) ⫽ .46. Second, age of child did not relate to change in the Aggression subscale measure from Time 1 to Time 2 (␤ ⫽ –.07). Youth internalizing problems. Child-reported depressive symptoms, assessed by the Child Depression Inventory (Kovacs, 1981), served as an indicator of internalizing problems. The Child Depression Inventory is a 27-item measure on which children indicate the degree to which they are experiencing depressive symptoms by using a scale ranging from 0 (You are sad once in a while) to 1 (You are sad many times) to 2 (e.g., You are sad all the time). Adequate reliability and validity data with samples similar to the one in this study have been reported (e.g., Fitzpatrick, 1993), and standardization data are available for children and adolescents ranging from 7 to 17 years old. The alpha coefficient for the current sample was .83. Scores can range from 0 to 81, with higher scores indicating greater depressive symptoms.

Procedure As previously mentioned, each assessment included two interviews: the sociodemographic interview and the psychological interview. The sociodemographic interviews were conducted by a team of five individuals (i.e., two PhD medical sociologists, one licensed social worker, two graduate students in public health), all with extensive experience working with inner-city African American families. All data for the current study were from the psychological interviews, which were conducted by a second team of 11 individuals (i.e., one PhD-level licensed clinical psychologist and ten clinical psychology doctoral candidate graduate students; eight women and three men, all Caucasians). All of these individuals had extensive training and experience with psychological interviewing and assessment. All interviewers were thoroughly trained in the use of their respective instruments during the pilot phases of the project, and interviewer training ensured cross-interviewer reliability and enhanced sensitivity to cultural and socioeconomic status differences. Interviewers were aware of HIV status of mothers but were not aware of the specific hypotheses of the current study. Each interviewer interviewed an approximately equal percentage of HIV-infected and non-infected mothers and children of these mothers. Each mother– child dyad in the non-infected sample was

186

JONES ET AL.

assessed at the child’s school, while each mother– child dyad in the maternal HIV/AIDS sample was assessed in a medical setting. When necessary, a taxicab was provided for transportation. On arrival to the first interview, the sociodemographic interview, the mother read and signed a consent form on behalf of herself and her child, and both were reassured of confidentiality. Subsequently, the mother and child were separated and privately administered questions assessing a range of sociodemographic information. The sociodemographic interview lasted approximately 1 hour. Families received $50 as compensation. The psychological interview was completed 2 days to 2 weeks after completion of the sociodemographic interview. As in the sociodemographic interview, the mother and child were separated, and each was interviewed separately to ensure privacy and confidentiality. The psychological interview lasted approximately 2 hours, and families again received $50 as compensation. For both the maternal HIV/AIDS and non-infected samples, all interview measures were administered verbally. In addition, cue cards were used in the psychological interview. These cue cards contained descriptors (i.e., “not true,” “sometimes true,” “often true”), along with corresponding numeric values (i.e., 0, 1, or 2) and pictorial representations (i.e., thermometers with various portions shaded). The same procedures were repeated 12 months later at the Time 2 assessment. All study and recruitment procedures were approved by the Institutional Review Boards at Louisiana State University Medical School, Tulane University, and the University of Georgia.

Results Preliminary Analyses The parent– child dyads who dropped out between Time 1 and Time 2 (n ⫽ 51) were compared with those who were retained across the two time points (n ⫽ 196) on demographic, independent, and dependent variables. No statistically significant differences emerged. Given our interest in the change in the child’s behavior problems over time, partial correlations were conducted

examining the associations between demographic variables and each outcome of interest (aggressive behavior and depressive symptoms at Time 2), controlling for the respective outcome variable at Time 1. Family income was associated with child depressive symptoms at Time 2 (r ⫽ –.14, p ⬍ .05), but no other demographic variables were associated with either aggressive behavior or depressive symptoms at Time 2. Thus, family income was statistically controlled only in multivariate analyses predicting depressive symptoms. In addition to the number of mothers who were HIVinfected, means, standard deviations, and range of scores for the major study variables and the correlations among these variables are presented in Table 2. On average, mothers reported relatively high levels of warmth and supervision, and children reported relatively low levels of aggressive behaviors and depressive symptoms. Furthermore, the two dependent variables (Time 2 aggressive behavior and depressive symptoms) had comparable standard deviations and ranges. Maternal HIV status was not associated with either parenting variable but was associated with child depressive symptoms at Time 1 (r ⫽ .19, p ⬍ .01) and child aggressive behavior at Time 2 (r ⫽ .17, p ⬍ .05). Mothers with HIV/AIDS had children who reported more depressive symptoms at Time 1 and aggressive behavior at Time 2. Given our interest in the moderating role of maternal HIV/ AIDS, as well as its associations with the outcomes of interest, maternal HIV status was entered in the first block of the regression analyses. As demonstrated in Table 2, maternal warmth was associated with depressive symptoms at Time 1 (r ⫽ –.19, p ⬍ .01) and with aggressive behavior at Time 1 (r ⫽ –.17, p ⬍ .05) and Time 2 (r ⫽ –.28, p ⬍ .01). Mothers who reported greater warmth in the mother– child relationship had children who reported fewer depressive symptoms and aggressive behaviors. Maternal supervision was associated with depressive symptoms at Time 1 (r ⫽ –.26, p ⬍ .01) and Time 2 (r ⫽ –.21, p ⬍ .05). Additionally, maternal supervision was associated with aggressive behavior at Time 1 (r ⫽ –.18,

Table 2 Descriptive Statistics for and Correlations Among Major Study Variables Variable

M

SD

Range

1. Number Mother HIV-infected (% HIV⫹; N ⫽ 71, 36%)a 2. Parental warmthb 3. Parental supervisionc 4. Depressive symptomsd (Time 1) 5. Depressive symptomsd (Time 2) 6. Aggressive behaviorse (Time1) 7. Aggressive behaviorse (Time 2)

14.75 50.36 9.40 7.44 6.27 5.74

4.53 5.97 6.47 5.96 5.63 5.44

1–19 17–56 0–42 0–29 0–26 0–26

—

p ⬍ .05. ** p ⬍ .01. 1 ⫽ HIV⫺ and 2 ⫽ HIV⫹ b Interaction Behavior Questionnaire, scale range 0 to 19. c Monitoring and Control Questionnaire, scale range 17 to 68. d Child Depression Inventory, scale range 0 to 81. e Youth Self-Report Aggression Subscale, scale range 0 to 57. * a

1

2

3

⫺.10 ⫺.12 — .18** —

4

5

.19** .09 ⫺.19** ⫺.13 ** ⫺.26 ⫺.21* — .54** —

6

7

.13 ⫺.17** ⫺.18** .47** .41** —

.17* ⫺.28** ⫺.12 .35** .56** .53** —


ns). Mothers who reported more supervision had children who reported fewer depressive symptoms and aggressive behaviors. Also of note, the two independent variables (maternal supervision and warmth) were significantly associated with one another at Time 1 (r ⫽ .18, p ⬍ .01), and the two dependent variables (depressive symptoms and aggressive behavior) were associated with one another at both time points (r ⫽ .47 and .53, respectively, p ⬍ .01). These findings point out the need to control for other parenting variables and child outcomes when examining parenting specificity.

Primary Analyses Hierarchical regression analyses were conducted to examine the proposed multivariate associations. In analyses predicting child depressive symptoms at Time 2, the following variables were entered in Block 1: maternal HIV status, family income, depressive symptoms at Time 1 (in order to control for existing depressive symptoms at Time 1), and aggressive behavior at Times 1 and 2 (in order to control for the covariation between outcome measures). The two parenting behaviors (i.e., maternal warmth and supervision) were entered in Block 2 (in order to examine each parenting behavior while controlling for the other). Analyses predicting child aggressive behavior at Time 2 were identical to those for child depressive symptoms with three exceptions: (a) family income did not need to be controlled via entry on Block 1; (b) aggressive symptoms at Time 1 (rather than depressive symptoms) were entered on Block 1 (in order to examine control for existing aggressive symptoms from Time 1 to Time 2); and (c) depressive symptoms at Time 1 and Time 2 (rather than aggressive behavior) were

187

entered on Block 1 (again, in order to control for the covariation between outcome measures). Analyses were also conducted with a third block examining the two-way interaction of maternal HIV status ⫻ each parenting variable as well as the two-way interaction of parental warmth ⫻ parental supervision. Maternal HIV status did not moderate the findings, and the two parenting variables did not interact; therefore, interaction terms were not considered further. Results from the regression analyses, reported in Table 3, indicate that parental warmth (␤ ⫽ –.16, p ⬍ .01), but not supervision (␤ ⫽ .05, ns), predicted change in child aggressive behavior from Time 1 to Time 2 (see top panel of Table 3). Higher levels of parental warmth at Time 1 predicted larger decreases/smaller increases in aggressive behavior from Time 1 to Time 2. In contrast, neither parental warmth (␤ ⫽ .08, ns) nor supervision, (␤ ⫽ –.06, ns) predicted change in child depressive symptoms from Time 1 to Time 2 (see bottom panel of Table 3). With regard to specificity, this pattern of findings suggests the following: (a) parental warmth, but not supervision, was associated with changes in aggressive behavior; and (b) parental warmth appeared to be differentially associated with changes in depressive symptoms and aggressive behavior from Time 1 to Time 2. In order to test the specificity of these two parenting behaviors with the two child outcomes, tests of difference between standardized beta coefficients were conducted based on procedures recommended by Cohen, Cohen, West, and Aiken (2003). Two sets of analyses were conducted. First, the standardized beta weights for supervision and for warmth were compared for analyses predicting aggressive behavior and for analyses predicting depressive symptoms. These com-

Table 3 Hierarchical Regression Analyses Predicting Change in Child Outcomes From Time 1 to Time 2 Block

F

⌬R2

␤

SE

t

Time 2 Aggressive Behavior 1. Block 1: Covariates Maternal HIV status Aggressive behavior (Time 1) Depressive symptoms (Time 1) Depressive symptoms (Time 2) 2. Block 2: Parenting variables Maternal warmth Maternal supervision

36.10**

26.33**

.43

.02

⫺.09 .36 ⫺.08 .45

.31 .06 .06 .06

⫺1.68 5.69** ⫺1.17 6.84

⫺.16 .05

.07 .05

⫺2.87** .95

.06 ⫺.13 .39 ⫺.01 .43

.34 .001 .06 .07 .07

1.21 ⫺2.44* 6.50** ⫺.01 6.93**

.08 ⫺.06

.07 .06

1.52 ⫺1.17

Time 2 Depressive Symptoms 1. Block 1: Covariates Maternal HIV status Family income Depressive symptoms (Time 1) Aggressive behavior (Time 1) Aggressive behavior (Time 2) 2. Block 2: Parenting variables Maternal warmth Maternal supervision *

p ⬍ .05.

**

p ⬍ .01.

33.96**

24.90**

.47

.01

188

JONES ET AL.

parisons afforded the opportunity to determine if one parenting behavior had a stronger unique association with each type of problem behavior than did the other parenting behavior. The findings from this first set of analyses indicated that the association of warmth and change from Time 1 to Time 2 in aggressive behavior (␤ ⫽ –.16, SE ⫽ .07) was significantly different than the association between supervision and change in aggressive behavior (␤ ⫽ .05, SE ⫽ .05), t(190) ⫽ 2.54, p ⬍ .05. The difference between the association of warmth and change from Time 1 to Time 2 in depressive symptoms (␤ ⫽ .08, SE ⫽ .07) and the association of supervision and change in depressive symptoms (␤ ⫽ –.06, SE ⫽ .06) was not significant at the p ⬍ .05 level, t(188) ⫽ 1.80. These findings provide some support for parenting specificity, as warmth and supervision differed in the strength of their association with aggressive behavior. The second set of analyses consisted of a comparison of the standardized beta weights for supervision when depressive symptoms versus aggressive behavior served as the dependent variable and for warmth when depressive symptoms versus aggressive behavior served as the dependent variable. These comparisons provided the opportunity to determine if each parenting behavior was differentially associated with aggressive behavior and depressive symptoms. For these comparisons, the predictor variables had to be identical across the two regression analyses conducted for aggressive behavior and depressive symptoms (Cohen et al., 2003). As a consequence, family income was added to the regression analysis when aggressive behavior served as the dependent variable; depressive symptoms at Time 2 were dropped from the regression analysis when change in aggressive behavior served as the dependent variable; and aggressive behavior at Time 2 was dropped from the regression analysis when change in depressive symptoms served as the dependent variable. Therefore, the variables entered into both of these two regressions were Block 1 – HIV status, family income, Time 1 aggressive behavior, and Time 1 depressive symptoms; and Block 2 – supervision and warmth. This second set of analyses revealed that the association between supervision and change in aggressive behavior from Time 1 to Time 2 (␤ ⫽ .03, SE ⫽ .06, ns) did not differ significantly from the association between supervision and change in depressive symptoms from Time 1 to Time 2 (␤ ⫽ .09, SE ⫽ .06), t(189) ⫽ 1.29, ns. However, the association between warmth and change in aggressive behavior (␤ ⫽ –.17, SE ⫽ .06, p ⬍ .01) did differ significantly from the association between warmth and change in depressive symptoms (␤ ⫽ .20, SE ⫽ .06), t(189) ⫽ 2.86, p ⬍ .01. Warmth was more strongly associated with decreases in aggressive behavior than in depressive symptoms. Interestingly, warmth was associated with increases in depressive symptoms from Time 1 to Time 2. As with the prior set of analyses, these findings provide some support for the specificity of parenting behavior, as warmth, but not supervision, was differentially associated with change in aggressive behavior and depressive symptoms.

Secondary Analyses As noted above, higher levels of maternal warmth were unexpectedly positively associated with increases in depressive symptoms from Time 1 to Time 2. One hypothesis for this finding is that child depressive symptoms “lead to” increased maternal warmth rather than vice versa. In order to test this hypothesis, we conducted a regression analysis in which maternal warmth at Time 2 was entered as the dependent variable and in which Time 1 child depressive symptoms, as well as Time 1 family income, supervision, warmth, aggressive behavior, and maternal HIV status, were entered as independent variables. As hypothesized, Time 1 child depressive symptoms were related to change in maternal warmth from Time 1 to Time 2 (␤ ⫽ .17, SE ⫽ .06, p ⬍ .01), such that higher levels of child depressive symptoms predicted an increase in maternal warmth. Importantly, the magnitude of this association did not differ from that reported earlier for Time 1 maternal warmth predicting change in child depressive symptoms from Time 1 to Time 2 (␤ ⫽ .20).

Discussion This longitudinal study examined the specificity of two parenting behaviors and two child outcomes among an inner-city African American sample, as well as the moderating role of maternal HIV/AIDS. Our hypotheses regarding unique and differential specificity were partially supported: Relative to parental supervision, parental warmth was associated with change in aggressive behavior but not depressive symptoms; and parental warmth, but not parental supervision, was differentially associated with the two child outcome measures. Contrary to our hypotheses, parenting specificity was not moderated by maternal HIV/AIDS. Our findings regarding the specificity of parenting and child aggressive behavior are consistent with theory (Young, 1970; 1974) and with a growing empirical literature highlighting the central role of warmth and support in the mother– child relationship for youths residing in highrisk and disadvantaged neighborhoods (Dishion & McMahon, 1998; Kim-Cohen, Moffitt, & Caspi, 2004; Masten & Coatsworth, 1998). Although beyond the scope of the current study, one explanation that has been proposed to explain the inverse association between parental warmth and child externalizing problems focuses on the child’s ability to regulate arousal (Tronick, 1989). Parenting characterized by low levels of warmth, support, and involvement has been shown to interfere with a child’s capacity to modulate and regulate arousal; as a result, a child may be less capable of considering the consequences of his or her actions and refraining from aggressive behavior (Brody et al., 2002). Although plausible, this mechanism has not yet been considered or examined in the context of parenting specificity. An understanding of mechanisms in this context will require explanations not only of the negative association between parental warmth– child aggressive behavior but also of the weaker association of parental supervision– child aggressive behavior and the positive association of parental warmth–


child depressive symptoms. Replication of the current findings with a similar sample should occur and, if successful, then potential mechanisms should be tested. Our findings are also consistent with behavioral parent training programs. That is, supervision often has not been emphasized in parenting interventions for aggressive behavior for children in the same age range as those in our sample; on the other hand, warmth is included as a critical component of behavioral parent training for these types of problem behaviors regardless of the age of the child (McMahon et al., 2006). Importantly, our findings do not negate the importance of supervision, which has been emphasized by others (e.g., Brody et al., 2001; Dishion & McMahon, 1998), for school age African American youths residing in disadvantaged urban neighborhoods; rather, our findings suggest that supervision may not have specific effects on specific outcomes for youths in this age range. The intervention literature also provides some support for our finding that warmth was a stronger predictor of decreases in externalizing than internalizing behaviors. That is, in a recent review of the child intervention literature, Barmish and Kendall (2005) concluded that there is less evidence to suggest that it is critical to include parents in interventions targeting child internalizing problems (e.g., anxiety) than there is evidence to suggest that it is critical to include parents in interventions targeting externalizing problems. Although the findings about the relative roles of warmth and supervision are consistent with the literature (e.g., Masten & Coatsworth, 1998) and with some of our hypotheses, they are not consistent with those of Caron et al. (2006), who found support for specificity of behavioral control but not warmth. Several theoretical, methodological, and empirical considerations may help to explain the differences in findings in the two studies. First, Caron and colleagues measured behavioral control more broadly, while we measured one dimension of behavioral control: supervision. The broader domain of behaviors (e.g., stating consequences, suggesting alternative behavior) assessed by Caron et al. may have had specific associations with specific child outcomes, whereas the narrower range of behaviors included in supervision may not have such associations. In addition, Caron and colleagues employed a cross-sectional, rather than longitudinal, design. Specificity of parenting behaviors may be differentially associated with change in child outcome over time. Third, the current sample appeared to live under more stressful circumstances than did the Caron et al. sample. For example, the average annual family income was approximately $8,000 for families in the current study, while the average family income in the Caron and colleagues study was $20,000. Specificity of parent behavior to child outcome may differ across environmental conditions in which families reside. One reason for these differences is that the meaning and implications of parenting behaviors may differ for children living in different contexts (Hill, 1995; Ispa et al., 2004). Finally, although the specificity of parenting in the current study was not moderated by maternal HIV status, a significant percentage of our sample, but not the sample studied by Caron et al., faced this additional burden. This difference may have influenced the discrepant

189

findings. Taken together, the variability in findings between our study and Caron and colleagues’ study suggests that further research on parenting specificity is warranted. Findings should be considered with the sociocultural context of the family, as should the design and measurement methodology employed. The results of the regression analyses examining unique and differential associations of parenting with child outcomes yielded one major difference. When examining the unique associations between warmth and depressive symptoms and controlling for aggressive behavior at Time 2, warmth was positively, but not significantly (␤ ⫽ .08), associated with changes in depressive symptoms. However, when aggressive behavior was no longer controlled in the analysis examining differential associations, warmth was positively and significantly (␤ ⫽ .20, p ⬍ .01) associated with changes in depressive symptoms. Two aspects of these findings merit comment. First, the positive association of warmth and depressive symptoms (␤ ⫽ .08 and ␤ ⫽ .20), particularly the latter significant one, is surprising in light of most of the literature indicating that warmth is associated with lower levels of internalizing problems (Garber et al., 1997). However, as we have noted, the association between parenting and either depressive or anxious symptoms, the two primary indicators of internalizing problems, is far less studied and understood than the association between parenting and aggressive behavior (see Barmish & Kendall, 2005; Horowitz & Garber, 2006; McKee et al, 2007; McMahon et al., 2006), and a building literature has suggested that parents may have a more important role in interventions for childhood internalizing disorders than we have traditionally suspected (Diamond & Josephson, 2005). However, a review of several related literatures does offer some potential explanations for this association. One potential explanation for the positive association between warmth and internalizing problems is that child behavior (i.e., internalizing problems) “leads” to maternal behavior (i.e., warmth) as mothers may be more warm and supportive when their child expresses or displays symptoms congruent with depression. We tested this hypothesis and found that the association between Time 1 child depressive symptoms and changes in maternal warmth from Time 1 to Time 2 was comparable with the association between Time 1 maternal warmth and changes in child depressive symptoms from Time 1 to Time 2 (␤ ⫽ .17 and .20). This suggests that child internalizing problems are as likely to lead to changes in maternal warmth as vice versa. Consistent with transactional systems theories that characterize the bidirectional nature of the interactions between children and their environments, the most likely explanation for this finding is that children’s depressive symptoms lead mothers to ramp up their warm and supportive parenting behavior (Bronfenbrenner, 1977; Cummings et al., 2000; Sameroff & Chandler, 1975). Building on interpersonal theories of depression, depressed children may be more likely to doubt the sincerity of increases in maternal warmth and support (i.e., “reassurance”), which in turn may exacerbate, rather than ameliorate, their depressive symptoms (e.g., Joiner, 1999; Joiner, Metalsky, Gencoz, & Gencoz, 2001; Prinstein, Borelli, & Cheah, 2005).

190

JONES ET AL.

A second potential explanation is that our measurement of warmth did not differentiate between contingent and noncontingent warmth. Attachment theorists have long highlighted the importance of parents’ responses being contingent on the verbal and nonverbal signals that children and adolescents provide (Ainsworth, Blehar & Waters, 1978). In turn, contingent or responsive parenting is thought to enhance adjustment by providing children with the opportunity to see that their cognitive, emotional, and physical needs will be met in predictable and supportive ways (Baumrind, 1978; Grusec & Goodnow, 1994; Rocisanno, Slade, & Lynch, 1985). Accordingly, future research on parenting specificity should differentiate contingent and noncontingent warmth and should examine their relative contributions to internalizing problems. A third explanation evolves around the strong association (r ⫽ .56, see Table 2) between depressive symptoms and aggressive behavior and the stronger association of parental warmth with depressive symptoms when aggressive behavior is not, rather than when it is, controlled: These two findings suggest that aspects of aggressive behavior play an important role in the unexpected positive association between parental warmth and child depressive symptoms. The exact nature of the role of aggressive behavior in this association merits attention in future research. However, regardless of the role, the findings suggest that it is critical to control not only other parenting variables but also other child outcomes when examining specificity of relationships between a parenting behavior and a child outcome. We hypothesized that HIV infection may attenuate the unique and differential association among parenting behaviors and child outcomes. This hypothesis was not supported. The failure to find a moderating role of HIV infection may suggest that our specificity findings are robust across groups of children who vary on exposure to this family stressor. Alternately, our sample size may not have been sufficiently large to detect an interaction effect for change scores on our outcome measures across a 1-year interval. As with all research, the current findings must be interpreted in the context of the study’s limitations. First, the families in the current study represented a low-income, inner-city, and largely single-mother-headed subsample of African American families. Although African American families are disproportionately represented in statistics on poverty and single-mother homes (e.g., Duncan, BrooksGunn, & Klebanov, 1994; McLoyd, 1990; U.S. Census Bureau, 2005), the findings should be generalized with caution to other socioeconomic and racial groups. Second, the samples constituted by HIV-infected and non-infected mothers were recruited by different procedures, were assessed at different places, and, once contacted, the noninfected sample was self-selected. All of these aspects of the study could have influenced the findings. Third, all of the measures (e.g., supervision) in this study were based on interviews with parents and children. Future studies should also include observation and coding of parent– child interactions, which were included in the Caron and colleagues (2006) investigation, another factor which may have contributed, at least in part, to the discrepant findings between

the two studies. Fourth, we examined parenting specificity with regard to maternal parenting alone; a more thorough understanding of parenting specificity will require consideration of the parenting behaviors of the other adults and family members who may be involved in parenting school age African American youths (e.g., fathers, grandmothers, etc.). Fifth, we utilized a measure of child aggressive behavior that was not validated on children as young as some of those included in the current sample. Although age of child was not correlated with the measure, and neither the alpha coefficient nor the mean score on the aggressive subscale of the Youth Self-Report differed between the younger children (under 11.0 years) and older children (11.0 years and older), the findings should be replicated with another measure of child aggressive behavior that has been validated for children in our age range. Finally, we examined two domains of parenting behaviors, warmth and supervision. Future longitudinal studies with African American youths should consider the specificity of other parenting behaviors such as psychological control, which was examined in the study by Caron et al. (2006), as well as hostility. This study had a number of methodological strengths that merit attention as well. First, the use of longitudinal data provided the opportunity to examine Time 2 child outcomes, controlling for the same outcome at Time 1. In this way, conclusions more consistent with directionality could be drawn (Cole & Maxwell, 2003; Loeber & Farrington, 1994). In addition, different family members reported on parenting (i.e., the mother) and child outcomes (i.e., the child), which decreased the likelihood that the obtained associations were due to common reporter variance (O’Connor, 2002). Third, each parenting behavior (e.g., warmth) was statistically controlled in analyses examining the association of the other parenting behavior (e.g., supervision) with child outcomes. By doing so, common variance between the two parenting behaviors was removed, affording the opportunity to examine unique associations between a specific parenting behavior and a specific child outcome. Finally, each child outcome (e.g., depressive symptoms) was statistically controlled when examining the other child outcome (e.g., aggressive behavior). Again, this approach afforded the opportunity to examine unique associations by removing common method variance. In conclusion, the current findings provided support for the specificity of maternal warmth, but not maternal supervision. Although replication is necessary before definitive clinical implications can be drawn, the current findings suggest that prevention and intervention programs targeting aggressive behavior among inner-city African American youths will benefit from the inclusion of a parenting intervention component and, more specifically, a parenting intervention targeting parental warmth.

References Achenbach, T. M. (1991). Integrative guide for the 1991 CBCL/ 4 –18, YSR, and TRF profiles. Burlington: University of Vermont, Department of Psychiatry. Ainsworth, M. S., Blehar, M. C., & Waters, E. (1978). Patterns of

SPECIFICITY OF MATERNAL PARENTING BEHAVIOR attachment: A psychological study of the Strange Situation. Oxford, England: Erlbaum. Anderson, R. N., & Smith, B. L. (2003). Deaths: Leading causes for 2001. National Vital Statistics Reports, 52, 27–33. Antle, B. J., Wells, L. M., & Goldie, R. S. (2001). Challenges of parenting for families living with HIV/AIDS. Social Work, 46, 159 –169. Armistead, L., Forehand, R., Brody, G., & Maguen, S. (2002). Parenting and child psychosocial adjustment in single parent African American families: Is community context important? Behavior Therapy, 33, 361–375. Armistead, L., Klein, K., & Forehand, R. (1995). Parental physical illness and child functioning. Clinical Psychology Review, 15, 409 – 422. Barmish, A. J., & Kendall, P. C. (2005). Should parents be coclients in cognitive-behavioral therapy for anxious youth? Journal of Clinical Child and Adolescent Psychology, 34, 569 –581. Baumrind, D. (1978). Parental disciplinary patterns and social competence in children. Youth and Society, 9, 239 –276. Baumrind, D. (1991). Effective parenting during the early adolescent transition. In P. A. Cowan & E. M. Hetherington (Eds.), Family transitions (pp. 111–163). Hillsdale, NJ: Lawrence Erlbaum Associates. Brody, G. H., Dorsey, S., Forehand, R., & Armistead, L. (2002). Unique and protective contributions of parenting and classroom processes to the adjustment of African American children living in single-parent families. Child Development, 73, 274 –286. Brody, G. H., Ge, X., Conger, R., Gibbons, F. X., McBride Murry, V., Gerrard, M., et al. (2001). The influence of neighborhood disadvantage, collective socialization, and parenting on African American children’s affiliation with deviant peers. Child Development, 72, 1231–1246. Brody, G. H., Murry, V. M., Gerrard, M., Gibbons, F. X., McNair, L., Brown, A. C., et al. (2006). African American Families Program: Prevention of youths’ high-risk behavior and a test of a model of change. Journal of Family Psychology, 20, 1–11. Bronfenbrenner, U. (1977). Toward an experimental ecology of human development. American Psychologist, 32, 513–531. Caron, A., Weiss, B., Harris, V., & Catron, T. (2006). Parenting behavior dimensions and child psychopathology: Specificity, task dependency, and interactive relations. Journal of Clinical Child and Adolescent Psychology, 35, 34 – 45. Centers for Disease Control and Prevention. (1992). 1993 revised classification system for HIV infection and expanded surveillance case definition for AIDS among adolescents and adults. Morbidity and Mortality Weekly Report, 41, 15. Centers for Disease Control and Prevention. (2004a). Diagnoses of HIV/AIDS—32 states, 2000 –2003. Morbidity and Mortality Weekly Report, 53, 1106 –1110. Centers for Disease Control and Prevention. (2004b). HIV/AIDS surveillance report, 2003. Atlanta, GA: U.S. Department of Health and Human Services. Chilcoat, H. D., Breslau, N., & Anthony, J. C. (1996). Potential barriers to parent monitoring: Social disadvantage, marital status, and maternal psychiatric disorder. Journal of the American Academy of Child & Adolescent Psychiatry, 35, 1673–1682. Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). Applied multiple regression/correlation analysis for the behavioral sciences. Mahwah, NJ: Lawrence Erlbaum Associates. Cole, D. A., & Maxwell, S. E. (2003). Testing mediational models with longitudinal data: Questions and tips in the use of structural equation modeling. Journal of Abnormal Psychology, 112, 558 – 577. Cummings, E. M., Davies, P. T., & Campbell, S. B. (2000).

191

Developmental psychopathology and family process: Theory, research, and clinical implications. New York: Guilford. Davidow, M., & Grusec, J. E. (2006). Untangling the links of parental responsiveness to distress and warmth to child outcomes. Child Development, 77, 44 –58. Diamond, G., & Josephson, A. (2005). Family-based treatment research: A 10 year update. Journal of the American Academy of Child & Adolescent Psychiatry, 44, 872– 887. Dishion, T. J., & McMahon, R. J. (1998). Parental monitoring and the prevention of child and adolescent problem behavior: A conceptual and empirical foundation. Clinical Child and Family Psychology Review, 1, 61–75. Dishion, T. J., Patterson, G. R., Stoolmiller, M., & Skinner, M. L. (1991). Family, school, and behavioral antecedents to early adolescent involvement with antisocial peers. Developmental Psychology, 27, 172–180. Duncan, G. J., Brooks-Gunn, J., & Klebanov, P. K. (1994). Economic deprivation and early childhood development. Child Development, 65, 296 –318. Dutra, R., Forehand, R., & Armistead, L. (2000). Child resiliency in inner-city families affected by HIV: The role of family variables. Behaviour Research and Therapy, 38, 471– 468. Family Health Project Research Group. (1998). A multidisciplinary longitudinal investigation of children whose mothers are HIV infected. Clinical Psychology Review, 18, 839 – 856. Fitzpatrick, K. M. (1993). Exposure to violence and presence of depression among low income African American youth. Journal of Consulting and Clinical Psychology, 61, 528 –531. Forehand, R., Miller, K. S., Armistead, L., Kotchick, B. A., & Long, N. (2004). The Parents Matter! Program: An introduction. Journal of Child and Family Studies, 13, 1–3. Forehand, R., Steele, R., Armistead, L., Morse, E., Simon, P., & Clark, L. (1998). The Family Health Project: Psychosocial adjustment of children whose mothers are HIV-infected. Journal of Consulting and Clinical Psychology, 66, 513–520. Garber, J., Robinson, N. S., & Valentiner, D. (1997). The relation between parenting and adolescent depression: Self-worth as a mediator. Journal of Adolescent Research, 12, 12–33. Gorman-Smith, D., & Tolan, P. (1998). The role of exposure to community violence and developmental problems among innercity youth. Development and Psychopathology, 10, 101–116. Granic, I., & Patterson, G. R. (2006). Toward a comprehensive model of antisocial development: A dynamic systems approach. Psychological Review, 113, 101–131. Gray, M. R., & Steinberg, L. (1999). Unpacking authoritative parenting: Reassessing a multidimensional construct. Journal of Marriage & the Family, 61, 574 –587. Grusec, J. E., & Goodnow, J. J. (1994). Impact of parental discipline methods on the child’s internalization of values: A reconceptualization of current points of view. Developmental Psychology, 30, 1–19. Hammen, C., Shih, J. H., & Brennan, P. A. (2004). Intergenerational transmission of depression: Test of an interpersonal stress model in a community sample. Journal of Consulting and Clinical Psychology, 72, 511–522. Hill, N. E. (1995). The relationship between family environment and parenting style: A preliminary study of African American families. Journal of Black Psychology, 21, 408 – 423. Horowitz, J. L., & Garber, J. (2006). The prevention of depressive symptoms in children and adolescents: A meta-analytic review. Journal of Consulting and Clinical Psychology, 74, 401– 415. Ispa, J. M., Fine, M. A., Halgunseth, L. C., Harper, S., Robinson, J., Boyce, L., et al. (2004). Maternal intrusiveness, maternal warmth, and mother–toddler relationship outcomes: Variations

192

JONES ET AL.

across low income ethnic and acculturation groups. Child Development, 75, 1613–1631. Joiner, T. E., Jr. (1999). A test of the interpersonal theory of depression in youth psychiatric inpatients. Journal of Abnormal Child Psychology, 27, 77– 85. Joiner, T. E., Jr., Metalsky, G. I., Gencoz, F., & Gencoz, T. (2001). The relative specificity of reassurance-seeking to depressive symptoms and diagnoses among clinical samples of adult and youth. Journal of Psychopathology and Behavioral Assessment, 22, 35– 41. Kim-Cohen, J., Moffitt, T. E., & Caspi, A. (2004). Genetic and environmental processes in young children’s resilience and vulnerability to socioeconomic deprivation. Child Development, 75, 651– 668. Kotchick, B. A., Forehand, R., Brody, G., Armistead, L., Simon, P., Morse, E., et al. (1997). The impact of maternal HIV infection on parenting in inner-city African American families. Journal of Family Psychology, 11, 447– 461. Kovacs, M. (1981). Rating scales to assess depression in schoolaged children. Acta Paedopsychiatrica: International Journal of Child & Adolescent Psychiatry, 46, 305–315. Lee, C. M., & Gotlib, I. H. (1991). Maternal depression and child adjustment: A longitudinal analysis. Journal of Abnormal Psychology, 98, 78 – 85. Loeber, R., & Farrington, D. P. (1994). Problems and solutions in longitudinal and experimental treatment studies of child psychopathology and delinquency. Journal of Consulting and Clinical Psychology, 62, 887–900. Masten, A. S., & Coatsworth, J. D. (1998). The development of competence in favorable and unfavorable environments: Lessons from research on successful children. American Psychologist, 53, 205–220. McKee, L., Colletti, C., Rakow, A., Jones, D. J., & Forehand, R. (2007). Parenting and child externalizing problems: Are the associations specific or diffuse? Manuscript submitted for publication. McLoyd, V. C. (1990). The impact of economic hardship on Black families and children: Psychological distress, parenting, and socioemotional development. Child Development, 61, 311–346. McMahon, R. J., Wells, K. C., & Kotler, J. S. (2006). Conduct problems. In E. J. Mash & R. A. Barkley (Eds.), Treatment of childhood disorders (pp. 137–268). New York: Guilford. O’Connor, T. G. (2002). The ‘effects’ of parenting reconsidered: Findings, challenges, and applications. Journal of Child Psychology and Psychiatry, 43, 555–572. Patterson, G. R., & Stouthamer-Loeber, M. (1984). The correlation of family management practices and delinquency. Child Development, 55, 1299 –1307. Prinstein, M. J., Borelli, J. L., & Cheah, C. S. L. (2005). Adolescent girls’ interpersonal vulnerability to depressive symptoms: A

longitudinal examination of reassurance-seeking and peer relationships. Journal of Abnormal Psychology, 114 [Special issue], 676 – 688. Prinz, R. J., Foster, S., Kent, R. N., & O’Leary, K. D. (1979). Multivariate assessment of conflict in distressed and nondistressed mother–adolescent dyads. Journal of Applied Behavior Analysis, 12, 691–700. Reid, M. J., Webster-Stratton, C., & Beauchaine, T. P. (2001). Parent training in Head Start: A comparison of program response among African American, Asian American, Caucasian, and Hispanic mothers. Prevention Science, 2, 209 –227. Rocissano, L., Slade, A., & Lynch, V. (1987). Dyadic synchrony and toddler compliance. Developmental Psychology, 23, 698 – 704. Sameroff, A., & Chandler, M. (1975). Reproductive risk and the continuum of caretaking causality. In F. Horowitz, M. Hetherington, S. Scarr-Salapatek, & G. Siegal (Eds.), Review of child development research (Vol. 4, pp. 187–244). Chicago: University of Chicago Press. Shaw, D. S., Winslow, E. B., Owens, E. B., Vondra, J. I., Cohn, J. F., & Bell, R. Q. (1998). The development of early externalizing problems among children from low-income families: A transformational perspective. Journal of Abnormal Child Psychology, 26, 95–107. Steinberg, L., Lamborn, S. D., Dornbusch, S. M., & Darling, N. (1992). Impact of parenting practices on adolescent achievement: Authoritative parenting, school involvement, and encouragement to succeed. Child Development, 63, 1266 –1281. Summers, P., Forehand, R., Armistead, L., & Tannenbaum, L. (1998). Parental divorce during early adolescence in Caucasian families: The role of family process variables in predicting the long term consequences for early adult psychosocial adjustment. Journal of Consulting and Clinical Psychology, 66, 327–336. Tompkins, T. L., Henker, B., & Whalen, C. K. (1999). Motherhood in the context of HIV/AIDS: Reading between the numbers. Cultural Diversity & Ethnic Minority Psychology, 5, 197–208. Tronick, E. Z. (1989). Emotions and emotional communication in infants. American Psychologist, 44, 112–119. U.S. Census Bureau. (2005, June 29). Current population survey, 2004 annual social and economic (ASEC) supplement. Retrieved October 1, 2005, from the U.S. Census Bureau Web site: http:// www.census.gov/population/www/socdemo/hh-fam/cps2004.html Young, V. H. (1970). Family and childhood in a Southern Negro Community. American Anthropologist, 72, 269 –288. Young, V. H. (1974). A Black American socialization pattern. American Ethnologist, 1, 415– 431.

Received November 7, 2006 Revision received September 24, 2007 Accepted September 27, 2007

䡲