Journal of Family Issues http://jfi.sagepub.com

Variations in Parenting and Adolescent Outcomes Among African American and Latino Families Living in Low-Income, Urban Areas Kathleen M. Roche, Margaret E. Ensminger and Andrew J. Cherlin Journal of Family Issues 2007; 28; 882 DOI: 10.1177/0192513X07299617 The online version of this article can be found at: http://jfi.sagepub.com/cgi/content/abstract/28/7/882

Published by: http://www.sagepublications.com

On behalf of: National Council on Family Relations

Additional services and information for Journal of Family Issues can be found at: Email Alerts: http://jfi.sagepub.com/cgi/alerts Subscriptions: http://jfi.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav Citations (this article cites 25 articles hosted on the SAGE Journals Online and HighWire Press platforms): http://jfi.sagepub.com/cgi/content/refs/28/7/882

Downloaded from http://jfi.sagepub.com at JOHNS HOPKINS UNIV on October 5, 2008

Variations in Parenting and Adolescent Outcomes Among African American and Latino Families Living in Low-Income, Urban Areas

Journal of Family Issues Volume 28 Number 7 July 2007 882-909 © 2007 Sage Publications 10.1177/0192513X07299617 http://jfi.sagepub.com hosted at http://online.sagepub.com

Kathleen M. Roche Margaret E. Ensminger Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland

Andrew J. Cherlin Johns Hopkins University, Baltimore, Maryland

Drawing from social disorganization theory, this study examined how perceived neighborhood conditions modified associations between parenting and delinquency, depressive symptoms, and school problem behavior among more than 800 African American and Latino 10- to 14-year-olds participating in Welfare, Children, and Families: A Three-City Study. Permissive and disengaged parenting, maternal involvement, and punitive parenting were associated with youth outcomes in varying ways depending on perceived neighborhood context and a youth’s race, ethnicity, and gender. Neighborhood-modifying influences on parenting were stronger for African Americans as compared to Latinos and for males as compared to females. Findings suggest that the stakes of uninvolved and permissive parenting for problematic youth outcomes are greater in higher risk neighborhoods. In addition, among African American males, punitive parenting is less strongly associated with poor youth outcomes when mothers perceive that the neighborhood poses more threats and offers fewer social resources. Keywords: African American and Latino families; parenting behaviors; youth behavior problems; low-income, urban neighborhoods

D

escriptions of the profound disadvantage characterizing many of this country’s poor, central-city neighborhoods have tended to overshadow the variability in family and adolescent functioning within these contexts. Among low-income, urban families, differences related to culture, gender,

882 Downloaded from http://jfi.sagepub.com at JOHNS HOPKINS UNIV on October 5, 2008

Roche et al. / Variations in Adolescent Outcomes

883

and perceived neighborhood conditions can influence the ways in which parenting is associated with successful adolescent development (GarciaColl & Pachter, 2002; Sampson, Raudenbush, & Earls, 1997). This study aims to advance our understanding of parenting among African American and Latino families who share in common the experience of being low income and living in poor or near poor, central-city neighborhoods. Among this population, we seek to understand how associations between parenting and adolescent problem behaviors vary by perceived neighborhood conditions and whether or not these neighborhood variations differ between males and females and between African Americans and Latinos.

Universal and Context-Specific Consequences of Parenting Generally, adolescents fare better when experiencing warm and supportive parenting, and they experience more problem behaviors when their parents are permissive or disengaged. Across racial and ethnic groups, adolescents report greater substance use and delinquency, poorer school performance, and worse psychological adjustment when parents are uninvolved or when discipline is lax or inconsistent. The negative consequences of permissive or disengaged parenting for adolescent well-being are in contrast to the benefits of warm, involved, and behaviorally regulating parenting (Pittman & Chase-Lansdale, 2001; Samaniego & Gonzales, 1999; Steinberg, Lamborn, Darling, Mounts, & Dornbusch, 1994; Steinberg, Mounts, Lamborn, & Dornbusch, 1991). Associations between parenting and adolescent outcomes are not uniform across contexts. In some studies, parental involvement in an adolescent’s schooling and homework has been associated with reduced youth problem behaviors and better academic outcomes only among White and socioeconomically advantaged families (Bogenschneider, 1997; McCreary & Dancy,

Authors’ Note: The authors of this study are grateful for Cheryl Alexander’s comments on an earlier version of this article. This study was carried out with core support from the National Institute of Child Health and Human Development and several other government agencies and private foundations. The development of the article was made possible by funding by the Johns Hopkins University (JHU) Center for the Prevention of Youth Violence (R49 CCR318627, P. Leaf, P. I.) and the JHU Center for Adolescent Health (U48 DP4000040, F. Sonenstein, P. I.), both of which are funded by the Centers for Disease Control and Prevention. Please address correspondence to Kathleen M. Roche, Department of Population, Family, and Reproductive Health, Johns Hopkins Bloomberg School of Public Health, 615 N. Wolfe Street, Room 4533, Baltimore, MD 21205; e-mail: [email protected].

Downloaded from http://jfi.sagepub.com at JOHNS HOPKINS UNIV on October 5, 2008

884

Journal of Family Issues

2004; McNeal, 1999). The extent to which parents allow their youth to make decisions related to free time and curfews (i.e., autonomy granting) has also shown important variations by the adolescent’s race, ethnicity, and gender. In one study, both the highest level of autonomy granting and the lowest level of autonomy granting were associated with greater depression among Asian adolescent females as compared to White, Latino, and African American adolescent females. Among boys, in contrast, the highest level of autonomy granting was more strongly associated with elevated depressive symptoms among African Americans as compared to Whites, Latinos, and Asians (Radziszewska, Richardson, Dent, & Flay, 1996). Although highly restrictive and physically punitive parenting has been associated with higher levels of emotional and behavioral problems across racial and ethnic groups (Eamon & Mulder, 2005; Grogan-Kaylor, 2005; McLeod & Nonnemaker, 2000), there has been some suggestion that the consequences of physical punishments are less apparent among African American compared to White youth. Research based on community samples, for example, has shown that corporal punishment is associated with problem behaviors among European American but not African American adolescents (Deater-Deckard & Dodge, 1997; Lansford, Deater-Deckard, Dodge, Bates, & Pettit, 2004). Furthermore, findings from research on preadolescent youth has shown modestly stronger associations between corporal punishment and youth behavior problems among White and Latino compared to African American youth (Eamon, 2001; McLeod & Nonnemaker, 2000). Among a national sample of youth aged 4 to 14, Grogan-Kaylor (2005) found that corporal punishment was more strongly associated with antisocial behaviors as youth age but that this age-related increase was less steep for African American and Latino as compared to White youth. This study also revealed that females’ antisocial behavior was less strongly associated with physical punishment when compared to that of males (Grogan-Kaylor, 2005). It is important to consider the social and economic context in which families live when exploring associations between parenting and youth outcomes. Young people living in poor, urban communities may perceive that high levels of restriction, behavioral control, and strict punishments are necessary given the importance of “emotional independence and physical survival” (Forehand & Kotchick, 1996, p. 193) in these contexts (Cauce, Stewart, Rodriguez, Cochran, & Ginzler, 2003; Furstenberg, Cook, Eccles, Elder, & Sameroff, 1999; Jarrett, 1997; Simons, Simons, & Wallace, 2004). Low levels of autonomy granting, parental restriction, and punitive parenting all have been shown to be less deleterious and in some cases more beneficial for adolescents living in higher risk communities as compared to those in lower risk

Downloaded from http://jfi.sagepub.com at JOHNS HOPKINS UNIV on October 5, 2008

Roche et al. / Variations in Adolescent Outcomes

885

communities (Baldwin, Baldwin, & Cole, 1990; Gonzales, Cauce, Friedman, & Mason, 1996). The extent to which neighborhood risk modifies parenting may vary by an adolescent’s racial and ethnic background. Among a sample of school-age children in Boston, parental restrictions were associated with better academic performance in more disadvantaged neighborhoods among African American and (to a lesser extent) Latino children. In that same study however, parental restrictions were associated with worse academic performance among White youth living in more disadvantaged neighborhoods (Dearing, 2004).

Parenting and Adolescent Outcomes Within a Neighborhood Context Social disorganization theory offers a useful framework for understanding why parenting’s effectiveness in deterring youth problem behaviors may vary depending on social attributes of the neighborhood context. Social disorganization refers to a community’s limited capacity to achieve consensus regarding residents’ values or to provide informal social controls for problematic behavior in a community. The social disorganization framework posits that economically fragile, residentially unstable neighborhoods discourage the formation of social ties among residents, resident involvement in community organizations, and informal social control over youth that in turn increase adolescent delinquent behaviors. Central to this theory is the idea that structurally disadvantaged communities result in greater delinquency and crime because of a breakdown in the community’s collective socialization of and informal control over youth (Sampson & Groves, 1989). Implicit within social disorganization theory is the idea that socially derived resources within a neighborhood represent an important source of social capital from which parents may draw in raising their adolescent youth (Sampson, 1992). A neighborhood’s social resources may also protect adolescents directly by helping buffer youth from the harmful consequences of ineffective parenting. For example, uninvolved and permissive parenting is likely to be less consequential for a youth’s involvement in problematic behavior in an environment where residents are willing to take actions aimed at preventing youth in the community from behaving in ways that threaten the safety and well-being of community residents (e.g., by vandalizing property, being loud and unruly in public settings, etc.). Although social disorganization is conceptualized as a neighborhood-level phenomenon, this study considers the relevance that individual-level perceptions of neighborhood social disorganization have to parent–youth relations.

Downloaded from http://jfi.sagepub.com at JOHNS HOPKINS UNIV on October 5, 2008

886

Journal of Family Issues

Thus, we consider variations in maternal perceptions of social cohesion and trust among neighbors and of neighbors’ willingness to collectively take responsibility for youth socialization and behavior. We also consider how mothers’ perceptions of neighborhood disorder and dangers vary by measuring problems such as perceived crime, drug dealing, gangs, and muggings. Based on the idea that uninvolved and lax parenting is more deleterious to adolescents living in threatening and socially disorganized neighborhoods, we expect that permissive and disengaged parenting styles and low maternal involvement in the adolescent’s schooling will be associated with greater increases in deleterious adolescent outcomes in dangerous and socially disorganized neighborhoods. Given the importance of behavioral restriction in higher risk environments, we anticipate that punitive parenting will be less strongly associated with adolescent problem behaviors in riskier neighborhood contexts. This study’s focus on a sample of low-income, urban African American and Latino families has the potential to inform our understanding of parenting among families who face considerable challenges in raising adolescents. Previous studies exploring neighborhood and parenting interactions have focused on either nationally representative samples of families (Roche et al., 2005), African American families in rural areas (Simons et al., 2002), or African Americans of diverse economic backgrounds (Gonzales et al., 1996). The present study’s use of longitudinal data and inclusion of Latino families improves on much of the previous literature (Furstenberg et al., 1999).

Data Source and Sample We use data from Welfare, Children, and Families: A Three-City Study (Cherlin, 1999), a study of low-income families in low-income Boston, Chicago, and San Antonio neighborhoods. Investigators collected data on demographic attributes, family processes, and health and well-being outcomes among caregivers and their children at two time points (in 1999 and again an average of 16 months later). A multistage, stratified, area probability sampling of dwelling units was used to identify clusters of census block groups from which eligible households were identified. A household was eligible if its income level was below 200% of poverty, female or couple headed, and had children under the age of 4 or ages 10 through 14. Families were subsampled at varying rates to control for income, race/ethnicity, family structure, and public assistance distributions. A total of 2,458 child and mother/female caregiver pairs responded to in-home, computer-assisted personal interviews

Downloaded from http://jfi.sagepub.com at JOHNS HOPKINS UNIV on October 5, 2008

Roche et al. / Variations in Adolescent Outcomes

887

(audio-computer assisted self-interviews were conducted for sensitive questions). About half (n = 1,147) of these pairs included youth aged 10 to 14. Because 90% of female caregivers were biological or adoptive mothers, we refer to them as mothers. Interviews of mothers with 10- to 14-year-olds took just more than 2 hours; youth interviews took just more than an hour. The study design and methods are described elsewhere (Winston, 1999). From the 1,147 mother–adolescent pairs, we selected the 1,023 youth who were either African American or Latino. The sample of White, non-Hispanic families was not large enough to permit separate analyses for Whites. Of the 1,023 youth, 9% had missing data for study variables and 11% did not complete a Time 2 interview. Final sample sizes were 830 (81%) for delinquency analyses, 831 (81%) for analyses related to school problems, and 823 (80%) for depression analyses. Youth excluded because of item nonresponse were younger, F(1, 1021) = 12.7, p < .001, and less likely to live in San Antonio, χ2(2) = 6.1, p < .05, than youth with complete data. Youth lost to attrition did not differ on study variables.

Measures Demographic variables. Using Time 1 data, we included measures of the youth’s gender, race/ethnicity, and age; family structure; and parent education. Gender was dummy coded with males as the omitted group. Age was measured in years. Using mother reports of the household roster, family structure included the following categories: mother and father/stepfather/male partner (the omitted category), mother only, mother and extended kin, and no biological or adoptive mother. Parent education indicated the highest level of education attained by either the mother or her spouse/male partner. Responses ranged from 1 = less than a high school education to 8 = graduate degree. Race/ethnicity was measured by adolescent self-report of being African American (the reference group) or Latino. Parenting. The Three-City Study’s items measuring parenting were developed based on both their relevance to major parenting constructs and their comprehensiveness in measuring the construct. The development of parenting scales was conducted with the guidance of leading psychologists with expertise in family functioning among Latino and African American families (Winston, 1999). For this study, we used principal components (PC) analyses to develop measures of parenting at Time 1. The five youth-reported items on punitive parenting submitted to PC analyses indicated youth reports of the mother scolding, spanking/hitting, and threatening to spank/hit; punish; and Downloaded from http://jfi.sagepub.com at JOHNS HOPKINS UNIV on October 5, 2008

888

Journal of Family Issues

threatening to kick her child out of the home. Four of the five items loaded at .60 or greater on a dimension of punitive parenting (the items on threatening to kick the child out of the home was removed because the factor loading was .49). Response categories included 1 = never in past 12 months, 2 = a few times, 3 = once a month or more, 4 = once a week or more, 5 = almost everyday (response categories fell along a 5-point Likert scale ranging from never true to always true). Maternal involvement in schooling was measured based on results from PC analyses indicating that three youth-reported items loaded similarly on a single factor. These items measured how often during the past 12 months the mother checked or asked about homework, talked about what the youth was learning in school or how the youth did on a test, and helped with homework or studying. Response categories included 1 = never, 2 = a few times, 3 = once a month or more, 4 = once a week or more, and 5 = almost everyday. The five mother-reported items submitted to PC analysis to develop a measure of permissive and disengaged parenting measured mother reports of avoiding giving chores, having rules, and dealing with the child; dropping rules if the child objects; and not having much time for the child (response categories fell along a 4-point Likert scale ranging from definitely false to definitely true, with higher scores reflecting greater permissive/disengaged parenting). The item on the mother not having much time for her child was dropped because its factor loading was .56. We used mean scores of summated punitive and permissive/disengaged items for youth with valid data on at least four of the five items comprising those scales and a mean score of involvement in schooling for youth with valid data for at least two of these three items. Parenting scales demonstrated moderate to good reliability (α = .70 for punitive, α = .75 for involvement, α = .64 for permissive/disengaged parenting). Subsequent to determining validity and reliability of parenting scales among the entire sample, we examined the psychometric properties of the parenting scales separately for African American and Latino youth. Results from PC analyses indicated that factor loadings for all three parenting scales were similar across racial and ethnic groups. In addition, there was no racial and ethnic difference in the reliability coefficient for punitive parenting. Cronbach’s alphas were only slightly higher among Latino families as compared to African American families for both maternal involvement in schooling (.64 vs. .59) and permissive/disengaged parenting (.76 vs. .74). Neighborhood conditions. Using cluster analysis of 16 mother-reported items on perceived neighborhood conditions, we developed a measure of four Downloaded from http://jfi.sagepub.com at JOHNS HOPKINS UNIV on October 5, 2008

Roche et al. / Variations in Adolescent Outcomes

889

different neighborhood typologies. Specifically, 4 neighborhood items measured social cohesion (e.g., “people in this neighborhood can be trusted,” “people around here are willing to help neighbors,” “this neighborhood is a good place to raise kids”), 4 items measured collective efficacy (e.g., “how likely is it that your neighbors would do something about children who were . . . spray-painting graffiti on a local building; skipping school and hanging out on a street corner; showing disrespect to an adult?”), and 8 items measured problems and dangers (e.g., “how much of a problem are . . . gangs in your neighborhood; unsupervised children in your neighborhood; assaults and muggings; abandoned houses?”). Likert-scale response categories ranged from strongly disagree to strongly agree (4 points) for social cohesion items, from very unlikely to very likely (5 points) for collective efficacy items, and from not a problem to a big problem (3 points) for neighborhood problems. Neighborhood items were derived from Sampson and colleagues’ (1997) 5-item Social Cohesion and 10-item Collective Efficacy scales and BrooksGunn, Duncan, Leventhal, and Aber’s (1997) item on neighborhood satisfaction. Results from the hierarchical cluster analysis using Ward’s method with a squared Euclidean distance revealed a theoretically meaningful four-cluster solution, which was confirmed by split-half analyses of random samples. Results from one-way analyses of variance with Bonferonni post hoc comparisons of means (not shown) revealed significant differences between clusters in regard to mothers’ perceptions of social cohesion, collective efficacy, and neighborhood problems. Distinctions between neighborhoods suggested the following typologies: Cluster 1—safe and socially organized (i.e., few problems, high collective efficacy, high social cohesion), Cluster 2— safe and individualistic (i.e., few problems, mid-high social cohesion, midlow collective efficacy), Cluster 3—dangerous and collectivist (i.e., many problems, mid-high collective efficacy, mid-low social cohesion), and Cluster 4—dangerous and socially disorganized (many problems, low efficacy, low cohesion). Supporting the validity of mothers’ reports of neighborhood dangers, interviewer reports of the environment immediately outside the home being unsafe (e.g., unlit entrance or stairway, broken steps, broken glass, alcohol, or drugs in entranceway or yard) were significantly associated with neighborhood typologies. Thus, 4% and 5% of families living in, respectively, safe and socially organized and safe and individualistic neighborhood types were rated by interviewers as having an unsafe environment, whereas, 10% and 14% of families in, respectively, dangerous and collective and dangerous and socially disorganized neighborhood typologies were observed to be living in an unsafe environment, χ2(3, 982) = 23.83, p < .001. Downloaded from http://jfi.sagepub.com at JOHNS HOPKINS UNIV on October 5, 2008

890

Journal of Family Issues

Adolescent problem behavior. We measured three kinds of youth problem behavior at Times 1 and 2: delinquency, school problems, and depression. Items for delinquency and school problems, which have been used with other samples of low-income racial and ethnic minority adolescents (Coley & Chase-Lansdale, 2000; Pittman & Chase-Lansdale, 2001), derived from the National Longitudinal Study of Youth (Borus, Carpenter, Crowley, & Daymont, 1982) and the Youth Deviance Scale (Gold, 1970; Steinberg et al., 1991). We used factor analysis and theory to guide the development of scales on delinquency and school problems. The six delinquency items indicated youth reports of stealing, getting in trouble with the police, damaging property, carrying a weapon, attacking someone, and fighting during the past year (α = .64, Time 1; α = .75, Time 2). The five items on school problems indicated youth reports of skipping school, copying homework, being suspended or expelled, getting detentions, and cheating (α = .57, Time 1; α = .63, Time 2). Response categories included 1 = never, 2 = once or twice, 3 = several times, and 4 = often. Depression was measured based on scores from the Brief Symptom Inventory 18, which has been shown to be a reliable and valid measure of depression (Derogatis, 2000). Response categories for the depression items fell along a 5-point Likert scale ranging from 1 = not at all to 5 = extremely (α = .75, Time 1; α = .83, Time 2). We used transformed scores for adolescent outcomes to address the skewed distribution of scores. For delinquency and school problems, transformed scales were based on the natural log of the mean of standardized items. For depression, a value of one was added to the raw score and the natural log of that score was used.

Analysis Plans This study’s biviarate analyses focused on variations in adolescent outcomes by parenting, neighborhood typology, race/ethnicity, and gender. In examining associations between parenting and adolescent outcomes, we examined Pearson correlation coefficients. We ran one-way analyses of variance (ANOVA) with F tests of significance to compare mean levels of parenting and adolescent outcomes by neighborhood context, gender, and race/ethnicity. Next, we ran main effects models for each of the three adolescent outcomes. We used lagged regression models where the adolescent’s Time 2 problem behavior was regressed on their Time 1 problem behavior, parenting and neighborhood variables, and demographic and family background attributes. Coefficients for independent variables in these models are interpreted as effects on changes in rates of adolescent problem behavior over time (Kessler & Greenberg, 1981). We ran additional main effects models Downloaded from http://jfi.sagepub.com at JOHNS HOPKINS UNIV on October 5, 2008

Roche et al. / Variations in Adolescent Outcomes

891

within gender and racial/ethnic subgroups to explore the extent to which neighborhood and parenting variables were associated with adolescent outcomes differently for boys and girls and for African Americans and Latinos. Because this study’s primary aim was to explore associations between parenting and adolescent outcomes by gender, race/ethnicity, and perceived neighborhood conditions, we tested the significance of several interaction terms. We began by including two-way interactions between each of the three parenting variables and neighborhood context within gender and race/ethnicstratified models. Interaction terms were entered in separate equations, and we retained those reaching statistical significance at a probability level less than .05. To asses whether or not two-way interactions between neighborhood typology and parenting differed significantly between racial/ethnic and gender groups, we ran additional models based on the entire sample that tested the significance of the following kinds of three-way interactions: Gender × Neighborhood Type × Parenting and Race/Ethnicity × Neighborhood Type × Parenting. We report on findings from interaction terms included in stratified models if the three-way interaction terms from the model using the entire sample confirmed the significance of differences between gender and racial/ethnic groups. We report interaction effects findings from models based on the entire sample when there are not statistically significant differences between gender or racial/ethnic groups. Some of the significant interaction effects are illustrated in figures. In these figures, we plot coefficients for changes in adolescent outcome associated with “low,” “mid,” and “high” levels of a particular parenting practice or style (i.e., at –1, 0, and +1 standard deviations of the parenting score) for each of the four neighborhood typologies.

Results Sample Descriptive Statistics As shown in Table 1, this study’s sample included slightly more girls than boys (53% vs. 47%), youth were an average of 12 years old, and just less than half of the youth (48%) were African American. More than two thirds of adolescents lived in families headed by a mother only (68%). Other youth were in families with a mother and kin (8%); a mother and father, stepfather, or male partner (16%); or no biological/adoptive mother (8%). The average level of parental education was a high school degree. Just less than one third of youth (31%) lived in neighborhoods characterized as safe and socially organized and one quarter lived in neighborhoods characterized as dangerous and socially disorganized. In all, 24% of youth lived in neighborhoods that

Downloaded from http://jfi.sagepub.com at JOHNS HOPKINS UNIV on October 5, 2008

892

Journal of Family Issues

Table 1 Parenting, Demographic Variables, Neighborhood Context, and Adolescent Outcomes: Descriptive Statistics Variable

%

Parenting Involved Permissive/disengaged Punitive Demographics Parent education Youth age Female Two parentsa Mother only Mother and kin No mother African American Neighborhood context San Antonio Boston Chicago Safe and socially organized Safe with moderately high social cohesion Dangerous with moderately high efficacy Dangerous and socially disorganized Youth problem behaviors Delinquency Time 2 Delinquency Time 1 School problems Time 2 School problems Time 1 Depression Time 2 Depression Time 1

M

SD

Range

3.81 1.54 2.02

1.10 0.63 0.81

1 to 5 1 to 4 1 to 5

1.02 12.0

0.56 1.39

0 to 3 10 to 14

–0.15 –0.14 –0.17 –0.17 0.89 0.86

0.47 0.46 0.55 0.53 0.88 0.84

–0.52 to 1.79 –0.59 to 1.47 –0.92 to 1.58 –0.83 to 1.52 0 to 3.22 0 to 3.22

α .75 .63 .70

53 16 68 8 8 48 32 35 33 31 24 20 25 .75 .64 .63 .57 .83 .75

Note: Because the individual Brief Symptom Inventory items are not available to the public, the reliability coefficients for depression scores are based on entire sample of 10- to 14-yearold study participants as reported by investigators of the Three-City Study (Winston, 1999). a. Two parents includes two biological/adoptive parents, mother and spouse, mother and male partner.

were safe and individualistic, and one fifth of youth lived in neighborhoods that were dangerous and collectivist. Youth were similarly proportioned in Boston, Chicago, and San Antonio. (Although we do not separately examine Latinos of different national origins or generational statuses because of the small sample sizes, it is useful to note that 50% of Latinos were of Mexican descent, 27% were Puerto Rican, 10% were Dominican, and 13% were from other national origins. In addition, 57% of Latinos were third-generation and

Downloaded from http://jfi.sagepub.com at JOHNS HOPKINS UNIV on October 5, 2008

Roche et al. / Variations in Adolescent Outcomes

893

38% were second-generation immigrants.) Descriptive statistics for demographic, parenting, neighborhood, and adolescent outcome variables are shown in Table 1. Correlations among parenting variables (not shown) indicated that maternal involvement in schooling was inversely associated with punitive parenting (r = –.18, p < .001) and to a lesser extent, with permissive and disengaged parenting (r = –.07, p < .05). Correlations among adolescent outcomes were stronger, ranging from .26 to .47 at Time 1 and from .31 to .51 at Time 2 (school problems and delinquent behavior were most strongly correlated). Youth experienced moderate changes in problem behavior and depressive symptoms over the roughly 16-month time period. This was indicated by correlations of .40 to .41 between Time 1 and Time 2 adolescent outcomes.

Parenting and Adolescent Outcomes: Bivariate Associations Bivariate correlations between parenting and adolescent outcomes indicated that higher levels of punitive parenting were associated with higher levels of depressive symptoms (r = .25, Time 1; r = .14, Time 2), delinquency (r = .20, Time 1; r = .16, Time 2), and school problems (r = .15, Time 1; r = .13, Time 2), all at a probability level of less than .001. Higher levels of maternal involvement in the adolescent’s schooling were also associated with lower levels of adolescent depressive symptoms at Time 1 (r = –.26, Time 1; r = –.21, Time 2, both at p < .001), delinquency (r = –.18, p < .001, Time 1; r = –.11, p < .01, Time 2), and school problems (r = –.26, Time 1; r = –.16, p < .001, Time 2). Permissive/disengaged parenting was significantly associated with more school problem behavior at Time 1 (r = .07, p < .05).

Neighborhood, Gender, and Racial/Ethnic Variations in Parenting and Adolescent Problem Behavior Table 2 presents results from one-way ANOVAs comparing mean scores for parenting and adolescent outcomes by neighborhood typology, race/ ethnicity, and gender. Overall, adolescents living in the two “safe” neighborhood clusters (i.e., safe and socially organized and safe and individualistic) reported less delinquency and fewer depressive symptoms at Time 2 when compared to adolescents living in the two “dangerous” neighborhood clusters, F(3, 813) = 3.2, p < .05. Among parenting variables, only maternal involvement in schooling varied by neighborhood type. Adolescents in safe and socially organized neighborhoods reported less maternal involvement in schooling than adolescents living in dangerous and socially disorganized

Downloaded from http://jfi.sagepub.com at JOHNS HOPKINS UNIV on October 5, 2008

894

Downloaded from http://jfi.sagepub.com at JOHNS HOPKINS UNIV on October 5, 2008

3.67a 1.51 1.82 –0.13 –0.16 –0.10 –0.13 0.87 0.84

3.73 1.51 1.84 –0.18 –0.21 –0.20 –0.23 0.80 0.78

2 n = 196 3.86 1.51 1.88 –0.09 –0.08 –0.14 –0.12 0.92 1.00

3 n = 160 3.96b 1.43 1.86 –0.13 –0.14 –0.21 –0.19 0.87 0.96

4 n = 203 3.89a 1.49 1.85 –0.07a –0.08a –0.15 –0.16 0.78 0.70

Male n = 383

Gender

3.71b 1.49 1.84 –0.19b –0.19b –0.16 –0.17 0.94 1.05

Female n = 430

3.92a 1.41a 2.07a –0.10a –0.07a –0.19 –0.21 0.83 0.87

African American n = 386

3.68b 1.66b 1.97b –0.17b –0.20b –0.13 –0.13 0.89 0.90

Latino n = 427

Race/Ethnicity

Note: Means in the same row not sharing the same subscript are significantly different based on results from one-way analyses of variance with an F test and Bonferroni tests of significance. Significance levels for race/ethnicity are p < .001 for Time 2 delinquency, p < .01 for permissive/disengaged and involved in schooling, and p < .001 for Time 1 delinquency and punitive. Significance levels for gender are p < .001 for Time 1 delinquency and Time 2 depression, p < .01 for Time 2 delinquency and Time 1 depression, and p < .05 for involved in schooling. Significance level for neighborhood clusters was p < .05. The F statistic indicated significant overall differences in Time 2 delinquency and Time 2 depression between neighborhoods at p < .05.

Involved in schooling Permissive/disengaged Punitive Delinquency Time 1 Delinquency Time 2 School problems Time 1 School problems Time 2 Depressive symptom Time 1 Depressive symptoms Time 2

1 n = 254

Neighborhood Typology

Table 2 Means of Parenting and Adolescent Outcomes by Neighborhood Typology, Adolescent Gender, Adolescent Race/Ethnicity

Roche et al. / Variations in Adolescent Outcomes

895

neighborhoods, F(3, 817) = 4.0, p < .05. When compared to Latino families, African American families reported higher levels of punitive parenting, F(1, 839) = 3.9, p < .05, and maternal involvement in schooling, F(1, 839) = 10.0, p < .01, and lower levels of permissive and disengaged parenting, F(1, 839) = 36.8, p < .001. In addition, African American youth reported greater delinquency than Latino adolescents at each time point, F(1, 839) = 4.4, p < .05 at Time 1 and F(1, 819) = 13.4, p < .001 at Time 2. Gender differences in parenting and adolescent outcomes were also manifest. Boys reported greater maternal involvement in schooling, F(1, 839) = 5.1, p < .05, less depressive symptomology, F(1, 839) = 7.7, p < .01 at Time 1 and F(1, 816) = 34.6, p < .001 at Time 2, and more delinquency, F(1, 839) = 14.9, p < .001 at Time 1 and F(1, 819) = 11.8, p < .01 at Time 2, when compared to girls.

Multivariate Results: Main Effects Models Results from main effects models for the entire sample are shown in Table 3. Across all outcomes, delinquency, school problem behavior, and depressive symptoms at Time 1 were significantly and strongly associated with the respective outcomes at Time 2 (B = .28 for delinquency, B = .33 for school problem behavior, and B = .34 for depressive symptoms, all at p < .001). Adolescent females reported greater decreases in delinquency (B = –.08, p < .05) and increases in depressive symptoms (B = .15, p < .001) than adolescent males. Although African Americans reported higher mean levels of delinquency at each time point, results from the lagged regression models indicate that Latinos reported greater increases in delinquency from Time 1 to Time 2 when compared to African Americans (B = .07, p < .05). In terms of parenting, punitive parenting (B = .08, p < .05) was associated with increased delinquency; punitive parenting was also associated with increased school problem behavior (B = .07, p < .05). Maternal involvement in schooling was associated with decreases in depressive symptoms (B = −.08, p < .05). Finally, living in a neighborhood characterized as dangerous and socially disorganized was associated with greater increases in depression than living in a safe and socially organized neighborhood. The variance explained for changes in delinquency was .23; in school problem behavior it was .19, and in depression it was .22. Results from main effects models stratified by gender and race/ethnicity indicated that parenting and neighborhood variables were associated with adolescent outcomes differently for boys, girls, African Americans, and Latinos. Among males, permissive and disengaged parenting was associated with increased delinquency (B = .12, p < .01) and school-related problem

Downloaded from http://jfi.sagepub.com at JOHNS HOPKINS UNIV on October 5, 2008

896

Downloaded from http://jfi.sagepub.com at JOHNS HOPKINS UNIV on October 5, 2008

.03 .01 .03 .02 .04 .04 .04 .03 .03 .19 14.0****

–.03 .06 .07 –.07 .07 .23

SE

.39 –.01 .05 .05

β

–.02 .05 .06 –.08** .08** .22

.38**** –.03 .06** .08**

B

–.06 .03 –.02 .02 –.06

.34 –.00 .05 .06

β

11.5****

.05 .05 .05 .04 .04

.04 .02 .03 .03

SE

–.05 .02 –.01 .02 –.05

.33**** –.00 .05 .07**

B

School Problem Behavior

–.00 .16 .16 .26 –.02

.36 –.06 –.01 .03

β

13.5****

.07 .08 .07 .06 .06

.03 .03 .05 .04

SE

Depression

–.00 .07 .08** .15**** –.01

.34**** –.08** –.01 .02

B

Note: Reference group: Neighborhood typology: safe and socially organized. Models include variables for age, parent structure, parent education, and city of residence. **p < .05.****p < .001.

Time 1 outcome Involved in schooling Permissive/disengaged parenting Punitive parenting Neighborhood typology Safe and socially organized Dangerous and collectivist Dangerous and socially disorganized Female gender Latino ethnicity R2 F

Variable

Delinquency

Table 3 Summary of Main Effects From Regression Analyses for Variables Predicting Adolescent Outcomes.

Roche et al. / Variations in Adolescent Outcomes

897

behavior (B = .12, p < .05), and living in a dangerous and socially disorganized neighborhood (as compared to a safe and socially organized neighborhood) was associated with increases in depressive symptoms. The only significant coefficient among females indicated that punitive parenting was associated with increased delinquency (B = .13, p < .01). Among African Americans, permissive and disengaged parenting (B = .11, p < .05) and punitive parenting (B = .12, p < .05) were associated with increased delinquency; punitive parenting was associated with increased school problems (B = .12, p < .05). Among Latinos, mothers’ involvement in schooling was associated with declines in delinquency (B = –.11, p < .05) and depressive symptomology (B = –.12, p < .01); living in a dangerous and collectivist neighborhood (as compared to a safe and socially organized neighborhood) was associated with increased delinquency (B = .12, p < .05).

Interaction Effects Models Results from models that examined interactions between parenting and neighborhood context are described separately for each parenting variable. We illustrate the significant interactions between parenting and neighborhood type for a subset of significant interactions. Coefficients from interaction effects models are included in the appendix.

Permissive and Disengaged Parenting × Neighborhood Context We begin by describing findings from interaction effects models that examine neighborhood variations in the association between permissive and disengaged parenting and adolescent outcomes. Among males, higher levels of permissive and disengaged parenting were associated with increased adolescent delinquency when families lived in dangerous or socially disorganized neighborhoods as compared to a safe and socially organized neighborhood. As shown in Figure 1, there is a positive slope for all neighborhood types except those characterized as safe and socially organized. The positive slope indicates that higher levels of permissive and disengaged parenting are associated with increases in delinquency scores over time. Although this interaction was significantly different by gender, there was no difference in the interaction between African American boys and Latino boys. Neighborhood-modifying influences on permissive and disengaged parenting were also apparent among African American males for school problems. Among African American males, permissive and disengaged parenting was most strongly associated with increased school problem behavior in dangerous and socially

Downloaded from http://jfi.sagepub.com at JOHNS HOPKINS UNIV on October 5, 2008

898

Journal of Family Issues

Figure 1 Change in Delinquency by Neighborhood Type and Permissive/Disengaged Parenting, Male Coefficient predicting changein delinquency

0 −0.1

Perceived Neighborhood Typology

−0.2

Safe and socially organized

−0.3

Safe and individualistic Dangerous and collectivist

−0.4

Dangerous and socially disorganized

−0.5 −0.6 −0.7 −1 SD

0

+1SD

Permissive/Disengaged Parenting

disorganized neighborhoods when compared to safe and socially organized neighborhoods. This interaction was significantly different from African American females and from Latino youth. Neighborhood conditions modified permissive and disengaged parenting among Latinos in different ways from that shown among African Americans. Permissive and disengaged parenting was associated with increased depression only in dangerous and collectivist neighborhoods. This style of parenting was actually associated with declines in Latino youths’ depressive symptoms in both dangerous and socially disorganized neighborhoods and safe and socially organized neighborhoods.

Punitive Parenting × Neighborhood Context Neighborhood variations in the association between punitive parenting and adolescent outcomes were mostly manifest among African American males. Punitive parenting was more strongly associated with declines in both delinquency and school problem behavior when African American males’ mothers perceived the neighborhood to be dangerous and socially disorganized or safe and individualistic. Punitive parenting was associated with increases in delinquency and school problem behavior for African American males in neighborhoods perceived as safe and socially organized or as dangerous and collectivistic. A third interaction between punitive

Downloaded from http://jfi.sagepub.com at JOHNS HOPKINS UNIV on October 5, 2008

Roche et al. / Variations in Adolescent Outcomes

899

Coefficient predicting changein depression

Figure 2 Change in Depression by Neighborhood Type and Punitive Parenting, African American Males 0.8 Perceived Neighborhood Typology

0.6

Safe and socially organized

0.4

Safe and individualistic 0.2

Dangerous and collectivist Dangerous and socially disorganized

0

−0.2 −0.4 −1 SD

0

+1 SD

Punitive Parenting

parenting and neighborhood context among African American males revealed that punitive parenting was associated with increases in depressive symptoms only within safe and socially organized neighborhoods and within dangerous and collectivist neighborhoods. Punitive parenting was not associated with changes in depressive symptoms in either safe and individualistic or dangerous and socially disorganized neighborhoods among African American males (Figure 2). The only other population for whom punitive parenting and neighborhood context were simultaneously related to adolescent outcomes was among Latino females. For these girls, punitive parenting was associated with sharp increases in depression only in the context of living in a dangerous and socially disorganized neighborhood.

Maternal Involvement in Schooling × Neighborhood Context Neighborhood-modifying influences on maternal involvement in schooling revealed consistent differences between African Americans and Latinos. Maternal involvement in schooling was associated with declines in delinquency, school-related problem behavior, and depressive symptoms only among African American youth living in dangerous and socially disorganized neighborhoods. The interaction for depressive symptoms is illustrated in

Downloaded from http://jfi.sagepub.com at JOHNS HOPKINS UNIV on October 5, 2008

900

Journal of Family Issues

Figure 3 Change in Depression by Neighborhood Type and Maternal Involvement in Schooling, African Americans

Coefficient predicting changein depression

0.8 Perceived Neighborhood Typology

0.6

Safe and socially organized

0.4

Safe and individualistic 0.2

Dangerous and collectivist Dangerous and socially disorganized

0

−0.2 −0.4 −1 SD

0

+1 SD

Maternal Involvement inSchooling

Figure 3. Findings related to maternal involvement in schooling and neighborhood context among Latinos differed from those found among African Americans. Among Latinos, maternal involvement in schooling was associated with declines in delinquency only among those living in dangerous and collectivist neighborhoods—that is, in neighborhoods with many threats, considerable resident willingness to intervene, and little social cohesion (Figure 4).

Discussion This study reinforces the idea that associations between parenting practices and youth behaviors are contextually specific rather than universal. Families in this study shared the common experience of being low income, representing racial and ethnic minority statuses, and living in low-income urban areas. Despite these shared experiences, differences related to a neighborhood’s physical and social conditions, youth gender, and youth race/ethnicity influenced associations between parenting and changes in adolescent delinquent behavior, school-related problem behavior, and depressive symptoms. Consistent with our hypothesis, there was strong evidence to suggest that the stakes of uninvolved parenting for adolescent problem behaviors are most profound when African Americans and males live in neighborhoods perceived as dangerous and socially isolating. This finding was supported by

Downloaded from http://jfi.sagepub.com at JOHNS HOPKINS UNIV on October 5, 2008

Roche et al. / Variations in Adolescent Outcomes

901

Figure 4 Change in Delinquency by Neighborhood Type and Maternal Involvement in Schooling, Latinos Coefficient predicting changein delinquency

0 −0.05

Perceived Neighborhood Typology

−0.1

Safe and socially organized −0.15 Safe and individualistic −0.2

Dangerous and collectivist

−0.25

Dangerous and socially disorganized

−0.3 −0.35 −0.4 −1 SD

0

+1 SD

Involvement in Schooling

five significant interaction terms and was manifest in three different ways. First, permissive and disengaged parenting was associated with increased delinquency among all males except those living in safe and socially organized neighborhoods. Permissive and disengaged parenting was most strongly associated with increased delinquency among males within dangerous and socially disorganized neighborhoods. Second, among African Americans, permissive and disengaged parenting was associated with increased schoolrelated problem behaviors only when youth lived in dangerous and socially disorganized neighborhoods. Third, lower levels of maternal involvement in their youth’s schooling was more strongly associated with increases in African American adolescents’ delinquency, school-related problem behavior, and depressive symptoms in dangerous and socially disorganized neighborhoods. Social disorganization theory emphasizes the idea that a neighborhood’s level of social organization facilitates positive parenting. Our findings suggest an alternative pathway such that neighborhood social capital and safety buffer youth from the negative consequences of permissive and uninvolved parenting. Adolescents living in more pernicious neighborhood contexts may have more exposure to and opportunities for engaging in deviant behaviors that, in turn, render them more vulnerable to the consequences of uninvolved parenting. As expected, neighborhood interactions related to punitive parenting differed from those found for permissive and disengaged parenting and maternal involvement in schooling. These were primarily manifest among African Downloaded from http://jfi.sagepub.com at JOHNS HOPKINS UNIV on October 5, 2008

902

Journal of Family Issues

American males only. Punitive parenting was less strongly associated with poor behavioral and psychological outcomes among African American adolescent males in dangerous and socially disorganized neighborhoods. We did not find that punitive parenting was related to worse adolescent outcomes only in highly socially organized neighborhoods. Rather, among African American males, punitive parenting was associated with increased depressive symptoms, delinquent behaviors, and school-related problem behaviors when youths’ mothers perceived living in safe and socially organized neighborhoods and in dangerous and collectivist neighborhoods. These two neighborhood types are distinguished from the others in that they are characterized by moderate to high levels of collective efficacy. The idea that parental behavioral restriction and punishment are less deleterious for youth living in higher risk communities validates findings from previous studies that have relied on relatively more economically diverse samples of youth (Baldwin et al., 1990; Gonzales et al., 1996). By demonstrating heterogeneity in parent–adolescent relations among a relatively homogeneous sample of low-income, urban racial and ethnic minority families, our findings underscore the sensitivity of parenting to perceived neighborhood conditions. Expanding on previous research, we have also shown that other aspects of parenting—namely, permissive and disengaged parenting and maternal involvement in schooling—have varying associations with youth outcomes depending on how mothers perceive the neighborhood context. The salience of collective efficacy to parenting and the behaviors and psychological well-being of African American males is consistent with the well-known studies from the Chicago Project on Human Development. In that research, Sampson and colleagues (1997) found that collective efficacy at a neighborhood level was central in explaining lower rates of crime in certain neighborhoods. Parents in this study who lived in high efficacy neighborhoods may have responded to their African American adolescent boys’ misbehavior with more punitive discipline because of the low individual and community tolerance for youth antisocial behavior. Although we did not have more than two time points of data in this study, information about parenting and youth behavior over time would help elucidate possible bidirectional pathways between youth behavior and parenting. Findings for African American males were in stark contrast to those found among Latino females for whom neighborhood risk appeared to magnify positive associations between punitive parenting and depressive symptoms. This study’s findings for the Latino adolescents and their families were more perplexing than those found among African Americans. Overall, Latino youth fared less well in neighborhoods characterized by high levels of crime and incivilities; relatively low levels of trust, support, and cohesion among Downloaded from http://jfi.sagepub.com at JOHNS HOPKINS UNIV on October 5, 2008

Roche et al. / Variations in Adolescent Outcomes

903

neighbors; and moderate to high levels of willingness among residents to intervene in youth deviant behavior. In addition, Latino youth did not experience increased problem behaviors when living in the most socially disorganized and dangerous neighborhoods as compared to safe and socially organized neighborhoods. Results for interactions between parenting and outcomes among Latino youth were consistent with these main effects findings. Only in a dangerous and collectivist neighborhood type was permissive and disengaged parenting associated with increased depression among Latino adolescents or was low maternal involvement in their youths’ schooling associated with increased delinquency. Results from post hoc analyses provided some clarification regarding the unexpected finding that Latino youth fared less well in dangerous and collectivist as compared to dangerous and socially disorganized neighborhoods. Results from one-way ANOVAs comparing mean scores of individual items comprising the neighborhood problems scale by neighborhood cluster indicated important differences between the two dangerous neighborhood clusters among Latinos. Specifically, Latino mothers perceiving the neighborhood to be dangerous and collectivist reported significantly greater problems with abandoned houses, drug dealing, and assaults and muggings than mothers perceiving the neighborhood to be dangerous and socially disorganized. (These distinctions were not apparent among African Americans.) This suggests that among Latinos, perceived resident willingness to intervene with unsupervised and deviant youth may be related to the dangers apparent in the neighborhood. That is, Latinos may be more likely to intervene in neighborhoods characterized by abandoned houses, drug dealing, and assaults, although these efforts are not necessarily successful. Differences related to parenting and neighborhood conditions that emerged between African Americans and Latinos in this study should be considered within the historical context of this country’s low-income, central-city neighborhoods. African Americans have resided in this country’s dangerous urban neighborhoods across more generations and for more years than have Latino families. As relative newcomers to poor, central-city neighborhoods in the United States, Latino families may perceive neighbors’ responses to community threats as more suspect than do African American families. The fact that our findings were much stronger among African Americans as compared to Latinos may also stem from the limited measures specific to Latino culture included in this study. We did not have measures of respect, obedience, and loyalty to family, which are all highly valued within Latino culture. In addition, our sample size was not large enough to permit separate analyses for Latinos of different national origins and generational statuses,

Downloaded from http://jfi.sagepub.com at JOHNS HOPKINS UNIV on October 5, 2008

904

Journal of Family Issues

both of which are relevant to parenting and adolescent outcomes (Harwood, Leyendecker, Carlson, Asencio, & Miller, 2002). Finally, although our focus on mothers’ parenting made sense given the small proportion of youth living with a father, this study would have been enhanced by a consideration of child rearing by other influential adults including fathers and other kin. The strengths of the study include the longitudinal data; the focus on information from low-income, central-city neighborhoods that vary in their characteristics; and having information from both parents and adolescents. Shared method variance related to the use of mother reports for both parenting and neighborhood conditions poses few problems to this study’s findings because the study’s dependent variable—youth outcomes—are reported by adolescents and also because mothers report on fairly concrete neighborhood conditions as opposed to feelings and emotions about living in the neighborhood. This study however is limited in some ways. First, we do not have more than two time points of data, which would enable us to explore bidirectional effects between youth behavior and parenting and between neighborhood attributes and parenting. It is likely that parents respond to both their environment and their youths’ behavior over time. Second, permissive and disengaged parenting and a punitive orientation toward parenting are likely to represent more enduring and intrinsic attributes of child rearing than some other practices such as monitoring and communication. In this way, our parenting variables may be less relevant to changes in adolescent behavior than to problem behaviors established prior to adolescence. This may explain the very small direct effects of parenting on changes in adolescent behavior. Examining specific practices such as monitoring, communication, and shared activities may yield stronger parent effects on trajectories of youth behavior because practices are more likely to change in response to the evolving developmental needs and behaviors of adolescents than are parenting styles. Researchers should also explore other social influences on adolescent behavior such as peer behaviors, the nature of the school context, and early adaptation to school, all of which are salient to adolescent behavior (Asteline, 1995; McNeely & Falci, 2004). This study’s results point to the importance of a neighborhood’s social and physical conditions for understanding associations between parenting and the behavior and adjustment of African American and Latino adolescents living in low-income, central-city neighborhoods. Knowledge about commonalities and deviations in parenting influences is of immense value for researchers interested in informing the development of preventive interventions focused on parenting. This is perhaps most important when considering families living in this nation’s most challenging neighborhood contexts.

Downloaded from http://jfi.sagepub.com at JOHNS HOPKINS UNIV on October 5, 2008

905

Downloaded from http://jfi.sagepub.com at JOHNS HOPKINS UNIV on October 5, 2008

All

Involved in schooling Permissive/disengaged –.07 Punitive parenting Neighborhoodc Safe, individualistic –.03 Dangerous, collectivist .04 Dangerous, disorganized .06 Permissive/Disengaged × Neighborhood Safe, Individualistic .09** Dangerous, collectivist .08* Dangerous, disorganized .11** Punitive × Neighborhood × Safe, Individualistic × Dangerous, Collectivist × Dangerous, Disorganized

Variable

–.08 .04 –.20**

–.02 .09* –.01 .14** .08 .24****

–.04 –.02 .03

–.05 .14 –.25**

.04 –.04 .05

–.03 –.01 .11

–.02 –.05 .12

–.17* .24*

.02

Latino

Depressive Symptoms

–.06 –.04 .01

.09

–.01 .05 .05

–.11

–.16* –.00 –.25**

–.02 .06 .07

–.06

(continued)

–.07 –.12** .02

–.07 .12 .02

.26*

African African African American African American Latino American Male Americana Latinob Male Female

School Problem Behavior

.19

.10

African American African Male American

Delinquency

Appendix Coefficients From Separate Regression Analyses for Variables Representing Interactions Between Neighborhood Typology and Parenting

906

Downloaded from http://jfi.sagepub.com at JOHNS HOPKINS UNIV on October 5, 2008

All .01 –.02 –.11†

–.04 –.19**** –.03

Latino

Depressive Symptoms

.00 –.01 –.15

African African African American African American Latino American Male Americana Latinob Male Female

School Problem Behavior

a. Race/Ethnicity × Neighborhood × Maternal Involvement significant; results for Neighborhood × Maternal Involvement not significant in model run among Latinos only. b. Race/Ethnicity × Neighborhood × Permissive and Disengaged significant; results for Neighborhood × Permissive and Disengaged not significant in model run among African Americans only. c. Neighborhood clusters are 1 = safe and socially organized (reference group), 2 = safe and individualistic, 3 = dangerous and collectivist, 4 = dangerous and socially disorganized. Models include all parenting and neighborhood variables, gender, Time 1 adolescent outcome, age, parent structure, parent education, and city of residence. *p < .10. **p < .05. ***p < .01.****p < .001.

Maternal Involvement × Neighborhood × Safe, Individualistic × Dangerous, Collectivist × Dangerous, Disorganized

Variable

African American African Male American

Delinquency

Appendix (continued)

Roche et al. / Variations in Adolescent Outcomes

907

References Asteline, R. H. (1995). A reconsideration of parental and peer influences on adolescent deviance. Journal of Health and Social Behavior, 36, 103-121. Baldwin, A. L., Baldwin, C., & Cole, R. E. (1990). Stress-resistant families and stressresistant children. In J. Rolf, A. S. Masten, D. Cicchetti, K. H. Nuechterlein, & S. Wintraub (Eds.), Risk and protective factors in the development of psychopathology (pp. 257-280). Cambridge, UK: Cambridge University Press. Bogenschneider, K. (1997). Parental involvement in adolescent schooling: A proximal process with transcontextual validity. Journal of Marriage and Family, 59, 718-733. Borus, M. E., Carpenter, S. W., Crowley, J. E., & Daymont, T. N. (1982). Pathways to the future, Volume II: A final report on the National Survey of Youth Labor Market Experience in 1980. Columbus: Center for Human Resource Research, Ohio State University. Brooks-Gunn, J., Duncan, G. J., Leventhal, T., & Aber, J. L. (1997). Lessons learned and future directions for research. In J. Brooks-Gunn, G. J. Duncan, & J. L. Aber (Eds.), Neighborhood poverty: Context and consequences for children (Vol. 1, pp. 279-297). New York: Russell Sage. Cauce, A. M., Stewart, A., Rodriguez, M. D., Cochran, B., & Ginzler, J. (2003). Overcoming the odds? Adolescent development in the context of urban poverty. In S. S. Luthar (Ed.), Resilience and vulnerability: Adaptation in the context of childhood adversities (pp. 343-363). Cambridge, UK: Cambridge University Press. Cherlin, A. J. (1999). Welfare, Children and Families: A Three-City Study, Wave 1 (Data Set 01-04) [Machine-readable data file and documentation]. Baltimore, MD: Johns Hopkins University. Coley, R. L., & Chase-Lansdale, P. L. (2000). Welfare receipt, financial strain, and AfricanAmerican adolescent functioning. Social Service Review, 74, 380-403. Derogatis, L. R. (2000). BSI 18: The Brief Symptom Inventory 18: Administration, scoring and procedures manual. Minneapolis, MN: National Computer Systems. Dearing, E. (2004). The developmental implications of restrictive and supportive parenting across neighborhoods and ethnicities: Exceptions are the rule. Applied Developmental Psychology, 25, 555-575. Deater-Deckard, K., & Dodge, K. A. (1997). Externalizing behavior problems and discipline revisited: Nonlinear effects and variation by culture, context, and gender. Psychological Inquiry, 8, 161-175. Eamon, M. K. (2001). Antecedents and socioemotional consequences of physical punishment on children in two-parent families. Child Abuse and Neglect, 6, 787-802. Eamon, M. K., & Mulder, C. (2005). Predicting antisocial behavior among Latino young adolescents: An ecological systems analysis. American Journal of Orthopsychiatry, 75, 117-127. Forehand, R., & Kotchick, B. A. (1996). Cultural diversity: A wake-up call for parents. Behavioral Therapy, 27, 187-206. Furstenberg, F. F., Cook, T. D., Eccles, J., Elder, G. H., & Sameroff, A. (1999). Managing to make it: Urban families and adolescent success. Chicago: University of Chicago Press. Garcia-Coll, C., & Pachter, L. M. (2002). Ethic and minority parenting. In M. H. Bornstein (Ed.), Handbook of parenting: Vol. 4. Social conditions and applied parenting (pp. 1-20). Mahwah, NJ: Lawrence Erlbaum. Gold, M. (1970). Delinquent behavior in an American city. Belmont, CA: Brooks/Cole. Gonzales, N. A., Cauce, A. M., Friedman, R. J., & Mason, C. A. (1996). Family, peer, and neighborhood influences on academic achievement among African-American adolescents: One-year prospective effects. American Journal of Community Psychology, 24, 365-387.

Downloaded from http://jfi.sagepub.com at JOHNS HOPKINS UNIV on October 5, 2008

908

Journal of Family Issues

Grogan-Kaylor, A. (2005). Corporal punishment and the growth trajectory of children’s antisocial behavior. Child Maltreatment, 10, 283-292. Harwood, R., Leyendecker, B., Carlson, V., Asencio, M., & Miller, A. (2002). Parenting among Latino families in the U. S. In M. H. Bornstein (Ed.), Handbook of parenting, Second edition, Volume 4: Social conditions and applied parenting (pp. 21-46). Mahwah, NJ: Lawrence Erlbaum. Jarrett, R. L. (1997). African American family and parenting strategies in impoverished neighborhoods. Qualitative Sociology, 20, 275-288. Kessler, R. C., & Greenberg, D. F. (1981). Linear panel analysis: Models of quantitative change. San Diego, CA: Academic Press. Lansford, J. E., Deater-Deckard, K., Dodge, K. A., Bates, J. E., & Pettit, G. S. (2004). Ethnic differences in the link between physical discipline and alter adolescent externalizing behaviors. Journal of Child Psychology and Psychiatry, 45, 801-812. McCreary, L. L., & Dancy, B. L. (2004). Dimensions of family functioning: Perspectives of low-income African American single-parent families. Journal of Marriage and Family, 66, 690-701. McLeod, J. D., & Nonnemaker, J. M. (2000). Poverty and child emotional and behavioral problems: Racial/ethnic differences in processes and effects. Journal of Health and Social Behavior, 41, 137-161. McNeely, C. A., & Falci, C. (2004). School connectedness and the transition into and out of health risk behavior among adolescents: A comparison of social belonging and teacher support. Journal of School Health, 74, 284-292. McNeal, R. B. (1999). Parental involvement as social capital: Differential effectiveness on science achievement, truancy, and dropping out. Social Forces, 78, 117-144. Pittman, L. D., & Chase-Lansdale, P. L. (2001). African American adolescent girls in impoverished communities: Parenting style and adolescent outcomes. Journal of Research on Adolescence, 11, 199-224. Radziszewska, B., Richardson, J. L., Dent, C. W., & Flay, B. R. (1996). Parenting style and adolescent depressive symptoms, smoking, and academic achievement: Ethnic, gender, and SES differences. Journal of Behavioral Medicine, 19, 289-305. Roche, K. M., Mekos, D., Alexander, C. S., Astone, N. M., Bandeen-Roche, K., & Ensminger, M. E. (2005). Parenting influences on early sex initiation among adolescents: How neighborhood matters. Journal of Family Issues, 26, 32-54. Samaniego, R. Y., & Gonzales, N. A. (1999). Multiple mediators of the effects of acculturation status on delinquency for Mexican American adolescents. American Journal of Community Psychology, 27, 189-210. Sampson, R. J. (1992). Family management and child development: Insights from social disorganization theory. In J. McCord (Ed.), Facts, framework, and forecasts: Advances in criminological theory 3 (pp. 63-93). New Brunswick, NJ: Transaction Publishers. Sampson, R. J., & Groves, W. B. (1989). Community structure and crime: Testing socialdisorganization theory. American Journal of Sociology, 94, 774-802. Sampson, R. J., Raudenbush, S., & Earls, F. (1997). Science, 277, 918-924. Simons, R. L., Lin, K., Gordon, L., Brody, G., Murray, V., & Conger, R. (2002). Community contextual differences in the effect of parental behavior on child conduct problems: A multilevel analysis with an African American sample. Journal of Marriage and Family, 64, 331-345. Simons, R. L., Simons, L. G., & Wallace, L. E. (2004). Families, delinquency, and crime: Linking society’s most fundamental social institution to antisocial behavior. Los Angeles: Roxbury.

Downloaded from http://jfi.sagepub.com at JOHNS HOPKINS UNIV on October 5, 2008

Roche et al. / Variations in Adolescent Outcomes

909

Steinberg, L., Lamborn, S. D., Darling, N., Mounts, N. S., & Dornbusch, S. M. (1994). Overtime changes in adjustment and competence among adolescents from authoritative, authoritarian, indulgent, and neglectful families. Child Development, 65, 754-770. Steinberg, L., Mounts, N., Lamborn, S., & Dornbusch, S. (1991). Authoritative parenting and adolescent adjustment across various ecological niches. Journal of Research on Adolescence, 1, 19-36. Winston, P. (1999). Welfare, Children and Families: A Three-City Study: Overview and design. Retrieved May 6, 2006, from http://www.jhu.edu/~welfare/overviewanddesign.pdf

Downloaded from http://jfi.sagepub.com at JOHNS HOPKINS UNIV on October 5, 2008