Youth & Society http://yas.sagepub.com

Early Entries Into Adult Roles: Associations With Aggressive Behavior From Early Adolescence Into Young Adulthood Kathleen M. Roche, Margaret E. Ensminger, Nicholas Ialongo, Jeanne M. Poduska and Sheppard G. Kellam Youth Society 2006; 38; 236 DOI: 10.1177/0044118X06287263 The online version of this article can be found at: http://yas.sagepub.com/cgi/content/abstract/38/2/236

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

Additional services and information for Youth & Society can be found at: Email Alerts: http://yas.sagepub.com/cgi/alerts Subscriptions: http://yas.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav

Downloaded from http://yas.sagepub.com at JOHNS HOPKINS UNIV on October 3, 2008

Early Entries Into Adult Roles

Youth & Society Volume 38 Number 2 December 2006 236-261 © 2006 Sage Publications 10.1177/0044118X06287263 http://yas.sagepub.com hosted at http://online.sagepub.com

Associations With Aggressive Behavior From Early Adolescence Into Young Adulthood Kathleen M. Roche Margaret E. Ensminger Nicholas Ialongo Johns Hopkins Bloomberg School of Public Health

Jeanne M. Poduska Sheppard G. Kellam American Institutes for Research

This study examines how early entries into adult roles are associated with aggressive and violent behavior occurring from early adolescence to young adulthood among 499 males and 578 females living in low-income, centralcity neighborhoods. Among males, engagement in adult roles accounted for the relationship between higher levels of aggressive behavior during early adolescence and violence in young adulthood. Males who lived independently and males who were behind in school or not in school at age 19 and 20 years (regardless of labor force participation) were at increased risk of engaging in violence in young adulthood. Among females, the positive association between early adolescent aggressive behavior and violence in young adulthood persisted even after accounting for involvement in adult roles. For females, parenthood and independent living were associated with involvement in violence during young adulthood in varying ways depending on labor force participation. Keywords: early adult roles; aggressive behavior; low income, urban youth

C

hildhood and adolescent aggression is an important risk factor for delinquent and violent behaviors occurring later in the life course (Broidy et al., 2003; Loeber & Hay, 1997; Maugham & Rutter, 1998; Moffitt, 1993). Research examining changes in antisocial and aggressive behaviors during adolescence commonly focuses on peer and parental influences 236 Downloaded from http://yas.sagepub.com at JOHNS HOPKINS UNIV on October 3, 2008

Roche et al. / Early Entries Into Adult Roles

237

(Moffitt, 1993; Patterson & Dishion, 1985; Patterson, Reid, & Dishion, 1992). In a departure from this research, we recently examined how a youth’s establishment of independence in domains of work, family (e.g., babysitting, doing chores), and peers (e.g., spending unsupervised time with peers) was associated with aggressive behavior among 12- and 13-year-olds who were low income living in poor, urban neighborhoods. Results from that published work (Roche, Ensminger, Chilcoat, & Storr, 2003) indicated that independent roles within work and peer domains were related to higher levels of aggression; whereas, familial independent roles were associated with less aggression. Because paid work modified associations that familial and peer independent roles had with aggression, we concluded that the meaning of a particular independent role depends on a youth’s entire repertoire of independent roles (Roche et al., 2003). Among this same population, the current study examines the association between independence and aggressive behavior from early adolescence through the transition to young adulthood. Rather than examining independence established during early adolescence, we focus on independence represented by adult roles such as parenthood, marriage, full-time work, and independent living. We explore this topic separately for males and females. When compared to females, males engage in more violence (Gove, 1985; Harris, 1977; Windle, 1990) and enter adult family roles (e.g., parenthood, marriage) earlier and extrafamilial adult roles (e.g., work) later (Chen & Kaplan, 1999; Michael & Tuma, 1985). Life-course perspective and social control theory (Hirschi, 1969) posit that early entries into adult roles deter individuals from continuing or becoming engaged in antisocial and risky behaviors. Marriage, parenthood, and employment are thought to confer benefits to individuals by providing informal social control over deviant behavior, strengthening investments and bonds in Authors’ Note: We would like to thank Scott Hubbard for assistance with data preparation. The development of the manuscript was made possible by funding by the Johns Hopkins University (JHU) Center for the Prevention of Youth Violence (R49 CCR318627, P. Leaf, primary investigator), funded by the Centers for Disease Control and Prevention. The data used for this article were derived form the Johns Hopkins Prevention Intervention Research Center Preventive Intervention Trials, which has been supported by grants from the National Institute of Mental Health (Epidemiologic Prevention Center for Early Risk Behaviors NIMH 5 PO MH38725, S. G. Kellam, primary investigator; Periodic Follow-up of Two Preventive Trials, MH42968, S. G. Kellam, primary investigator; Development & Malleability for Childhood to Adulthood, MH 42968-06A2, S. G. Kellam, primary investigator) and the National Institute on Drug Abuse (Transitions to Adulthood: The Baltimore Study, NIDA DA09887, J. C. Anthony, primary investigator; Etiology and prevention of drug related behaviors DA 04392). Correspondence regarding this article should be addressed to: Kathleen M. Roche, 615 N. Wolfe Street, E-4005, Baltimore, MD 21205, e-mail: [email protected].

Downloaded from http://yas.sagepub.com at JOHNS HOPKINS UNIV on October 3, 2008

238

Youth & Society

social institutions, and increasing individual responsibility (Horney, Osgood, & Marshall, 1995; Nielsen, 1999). Sampson and Laub (1990), for example, found that job stability and attachment to a spouse were associated with declines in antisocial behaviors throughout the life course, regardless of childhood delinquency. Similarly, in a longitudinal study of more than 600 males who had committed serious criminal offenses, males who reported living with a spouse experienced a 57% decrease in the odds of later committing an assault; whereas, those who lived with a girlfriend experienced a significant increase in criminal offending. In this same study, attending school was associated with declines in criminal offending and, contrary to authors’ expectations, engaging in paid work was not (Horney et al., 1995). The benefits of marriage may depend on one’s racial and ethnic background. Marriage has been associated with increased affluence and less problematic alcohol use among Whites and Latinos but not African Americans (Hirschl, Altobelli, & Rank, 2003; Nielsen, 1999). Research based on a life-span perspective emphasizes the importance of the timing of adult roles for deviant behavior (Nielsen, 1999). The focus in this regard has been on the negative social and economic consequences of precocious entries into independent living, parenthood, marriage, and full-time work (Aquilino, 1996; Buck & Scott, 1993; Card & Wise, 1978; Newcomb & Bentler, 1988; Uggen, 2000). Among a national sample of youth, living independently outside of marriage during late adolescence and young adulthood was related to greater increases in alcohol and marijuana use during the 3 years following high school (Bachman, O’Malley, & Johnston, 1984). In addition, early childbearing has been associated with lower labor force earnings and educational attainment and greater psychological distress and family disruption later in the life course among African Americans who are low income (Kellam, Adams, Brown, & Ensminger, 1982; Thompson & Ensminger, 1989). The theoretical base for the data collected here—the life course and social field framework—combines a social-role perspective with individual developmental histories (Kellam, 1990; Kellam, Branch, Agrawal, & Ensminger, 1975; Kellam & Ensminger, 1980; Kellam & Rebok, 1992). Across the life course, different social fields have varying prominence. For example, work and intimate social fields take on increasing importance as individuals traverse adolescence into adulthood. Within a life course and social field framework, normal development is marked by the integration of earlier competencies into later competencies. A major premise of this perspective is that social adaptation results from the interaction between the social field and rater characteristics and the characteristics of the individual. For example, teachers are the

Downloaded from http://yas.sagepub.com at JOHNS HOPKINS UNIV on October 3, 2008

Roche et al. / Early Entries Into Adult Roles

239

natural raters in the social field of the classroom. Their ratings have importance not only because they are observers of the child’s behavior but also because they are legally sanctioned to give grades and to pass or fail the student. This does not imply that the teachers are necessarily unbiased or correct in their observations but that they are important raters who have influence over the child’s trajectory. Early successful social adaptation in the face of developmental challenges promotes later adaptation as one encounters new social task demands (Cicchetti & Schneider-Rosen, 1984). Successful entries into age-appropriate social roles, such as being a student and having positive relationships with peers, provide competencies for later life success in work and parenting; whereas, early involvement in aggressive behavior can diminish the chances of later successful role performance. Although most young people in the United States do not enter parenthood or full-time work by the time they are age 19 and 20 years, youth living in communities and families with limited social, human, and financial capital commonly face the responsibilities of adulthood at a young age (Arnett, 2001; Burton, Allison, & Obeidallah, 1995; Marini, 1985). Family socioeconomic status has been positively associated with the age of labor force entry, first marriage, and first birth (Marini, 1985). Living in a stepparent or single-parent family also has been positively associated with leaving the home at a younger age (Aquilino, 1991; Buck & Scott, 1993; Cooney & Mortimer, 1999; Smith et al., 2000; Tang, 1997), marrying early (Michael & Tuma, 1985), bearing children early (Wu & Martinson, 1993), and entering the labor force prior to the completion of school (Chen & Kaplan, 1999). Many African American youth who are low income in the United States’ poor, urban neighborhoods do not perceive a good future and have substantial adult responsibilities. Many of these youth thus experience an “accelerated” adolescence (Burton et al., 1995). Building on our earlier study of independence and aggression among urban youth who are low income and on research examining adult roles, the current study examines how early entries into adult roles are relevant to a youth’s aggressive and violent behavior as it occurs from early adolescence through the transition to young adulthood. We consider the adult roles of independent living, parenthood, marriage, and full-time work. The current study’s first research question is based on the premise that an early entry into adult roles is a consequence of early problem behavior and the impetus for youth continuing or ceasing—depending on the kind of adult role—later involvement in problem behavior. We ask: “Is the association between early adolescent aggression and young adult violence accounted for by a youth’s entry into early adult roles?” We expect that early aggression will be associated with early entries into full-time work, marriage, independent

Downloaded from http://yas.sagepub.com at JOHNS HOPKINS UNIV on October 3, 2008

240

Youth & Society

living, and parenthood. Based on findings from our previous study that suggested protective effects of familial independent roles and harmful consequences of extrafamilial independent roles for aggressive behavior (Roche et al., 2003), we expect that employment and independent living will be associated with greater violence and that parenthood and marriage will be associated with less violence during young adulthood. Because our previous research revealed interactive influences of independent roles, we pose a second research question in the current study: “Is a single adult role associated with violence in young adulthood differently depending on one’s engagement in other adult roles?” We expect parenthood to offset some detrimental consequences of living independently for violence because the responsibilities of parenthood provide sanctions and social controls within the structure of independent living. We also expect full-time work to be less strongly associated with violence for youth who are still living with their family of origin because of the social controls provided by the family context. We stratify analyses by gender because of differences in rates of physical aggression and violence and in the timing of entry into adult roles among males and females. We account for the individual’s poverty status in grade school and her or his racial background because previous research has shown that youth who are poor and have a racial or ethnic minority background experience earlier entries into adult roles (Arnett, 2001; Marini, 1985) and more aggression (Centers for Disease Control and Prevention, 2002; Patterson et al., 1992).

Method Sample We use data from an epidemiologically defined study of students in the Baltimore City Public School System who participated in a longitudinal prevention study conducted by the Johns Hopkins University Prevention Research Center in partnership with the local public school system. The study included two successive cohorts of first graders. The first cohort enrolled in first grade in 1985, and the second cohort enrolled in 1986 (Dolan et al., 1993; Kellam, Werthamer-Larsson, Dolan, & Brown, 1991). Five different urban areas within one large elementary school district in eastern Baltimore were selected with the involvement of the Baltimore City Planning Department. These areas varied in terms of ethnicity, type of housing, family structure, socioeconomic status, violent crime, suicide and school drop-out rates. Three or four schools

Downloaded from http://yas.sagepub.com at JOHNS HOPKINS UNIV on October 3, 2008

Roche et al. / Early Entries Into Adult Roles

241

were selected in each urban area that were matched with regard to census tract, grade level, and first- and second-grade data; and, in first grade, schools, teachers, and children were randomly assigned to classrooms and to intervention or control statuses. Students who stayed in the Baltimore City Public School System and their teachers participated in assessments in the fall and spring of first and second grades, and in subsequent springs through middle school. All students were followed up and interviewed by telephone for 2 hours at ages 19 to 21 years. Of the 2,311 youth originally enrolled in Cohorts 1 and 2, the current study sample included 499 males and 578 females (47%, n = 1,077). This sample comprised youth who were assessed by a teacher in seventh grade (59%, n = 1,363) and who completed an interview at age 19 and 20 years (74%, n = 1,715). A small proportion (3%, n = 37,) of participants were excluded because they had missing data on study variables. Males were more likely than females, 32% vs. 19%; χ2(1) = 49.7, p < .001, and White, non-Hispanics were more likely than African Americans to not complete the young adult interview, 32% vs. 21%; χ2(1) = 61.6, p < .001. Those not completing the young adult interview were also reported by teachers in the fall of first grade to be more aggressive and disruptive, M = 2.0 versus M = 1.87; F(1, 1869) = 6.6, p < .05, and to have more concentration or attention problems, M = 3.02 vs. M = 1.76; F(1, 1869) = 11.3, p < .01, than others. White non-Hispanic youth were more likely than African American youth to be excluded from our sample based on not having been assessed by a teacher in seventh grade, 65% vs. 29%; χ2(1) = 271.1, p < .001. (The study assessed students who were still in the Baltimore City School System in seventh grade, and White students were more likely to have moved outside the city). Youth not assessed by a teacher in seventh grade were reported by teachers in the fall of first grade to be shyer than others, M = 2.75 vs. M = 2.63; F(1, 2,053) = 7.1, p < .01.

Measures Violence in emerging adulthood. At age 19 and 20, participants answered the following questions: “In the last year, have you” (a) used force to get money or expensive things from another person? (b) carried a hand gun when you went out? and (c) taken part in a fight? (response categories included no and yes). Among males, 9% (n = 43) had used force to steal, 19% (n = 93) had carried a gun, and 37% (n = 184) had taken part in a fight. Because almost one half of males (43%, n = 215) had engaged in one of these three forms of violence, our outcome measure of violence for males was indicated by either carrying a gun or using force to steal. Although gun Downloaded from http://yas.sagepub.com at JOHNS HOPKINS UNIV on October 3, 2008

242

Youth & Society

carrying and using force to steal are distinct forms of violence, each poses important risks to an individual’s health and safety. More than one fifth (22%, n = 112) of males reported serious violence during young adulthood. A small proportion of females reported using force to steal (3%, n = 18), carrying a gun (2%, n = 13), and being in a fight (13%, n = 74). Our measure of violence among females was, therefore, indicated by involvement in any of the three behaviors. Sixteen percent (n = 90) of the females reported violence. Adult roles. We measured four adult roles: parenthood, marriage, independent living, and work and/or schooling. Parenthood was indicated by responses to the question: How many children do you have? Those with at least one child were given a value of 1 for this measure, indicating that the respondent was a parent (the reference group was coded as 0). One third (n = 347) of 19- and 20-year-olds reported being a parent. Marriage was measured by the young adults’ reports of ever having been married (0 = no, 1 = yes). Only 2% (n = 19) of respondents were married. Independent living was measured by respondent reports of whose place they stayed most often during the past month. For respondents who did not report staying with a parent, grandparent, or aunt and/or uncle, a value of 1 was given for the independent living measure (0 was given to all others, the reference group). Just less than one half (45%, n = 481) of participants lived independently at age 19 and 20 years. To measure respondents’ work and/or schooling status, we combined responses from reported work hours and schooling status. Using responses to an item asking how many hours the respondent works in a typical week, we created an indicator of full-time work based on the respondent working at least 35 hours per week. Schooling was measured by a question regarding the participant’s current course of study. We grouped responses of high school, GED program, and not in school into a category indicating that the youth was not in school or behind in school; we grouped responses of university, 4-year program; community college, 2-year program; and vocational/business/ technical program into a category indicating that the youth was enrolled in advanced education. Although youth within the not in school/behind in school category may differ with respect to having a high school education, we group them together for the current study largely because a high school education does not buy youth better jobs in today’s economy. We then created the following work and/or schooling categories dummy coded as 0 or 1: not in school or behind in school and not working full-time (the reference group; 28%, n = 300); not in or behind in school and working full-time (42%, n = 450); enrolled in advanced education and not working full-time (17%, n = 183); and enrolled in advanced education and working full-time (13%, n = 142). Downloaded from http://yas.sagepub.com at JOHNS HOPKINS UNIV on October 3, 2008

Roche et al. / Early Entries Into Adult Roles

243

We used the Teacher Observation of Classroom Adaptation–Revised (TOCA-R; Werthamer-Larsson, Kellam, & Wheeler, 1991) to measure seventh grade aggressive behavior. The TOCA-R is a structured interview administered by a trained interviewer who follows a precise script; interviewers respond to teacher-initiated issues in a standardized fashion. Teachers first responded to items pertaining to the child’s adaptation in the classroom, they then made global ratings of each child’s behavior and progress. In these analyses, we use three domains on which teachers rate each child’s performance during the previous 3 weeks: aggressive behavior, social isolation, and concentration problems. A single teacher, usually the English/Language arts teacher, reported on the youth’s aggressive behavior; this teacher was unaware of the student’s first-grade intervention status. Teacher ratings of aggressive behavior among this sample consistently predicted official juvenile court records across teachers and across grades (Petras, Chilcoat, Leaf, Ialongo, & Kellam, 2004). This article uses the aggressive scale including the following items: (a) breaks rules, (b) harms others and property, (c) breaks things, (d) takes others property, (e) fights, (f) lies, (g) teases classmates, (h) yells at others, (i) stubborn (response categories range from 1 = never to 6 = almost always). Coefficient alphas ranged from .92 to .94 for Grades 1 through 7, or ages 8 to 13 years. Werthamer-Larsson et al. (1991) reported test-retest correlations during a 4-month interval with different interviewers of .60 or higher for the scale, and aggression ratings were correlated at .67 (p < .001) with peer ratings of aggression. Using gender-specific cut-points, we categorized scores into tertiles indicating 0 = low, 1 = moderate, and 2 = high aggression. Demographic characteristics. Race indicates that the respondent is White, non-Hispanic, as compared to African American, the reference group. Eighty percent (n = 395) of males and 83% of females (n = 480) were African American. We measure whether the youth is poor based on school records of the child receiving a free or reduced price lunch (the reference group is nonpoor). About two thirds of males (68%, n = 335) and females (66%, n = 384) in this sample were poor.

Analysis Plans We initially ran gender-stratified, cross-tabulations with chi-square tests of significance to examine how adult-role participation at age 19 and 20 years varied by a youth’s level of aggression during early adolescence and by violence during young adulthood. Next, we ran multivariate analyses aimed at answering each of the two research questions. Addressing the first research question, we ran two-step logistic regression models separately for males Downloaded from http://yas.sagepub.com at JOHNS HOPKINS UNIV on October 3, 2008

244

Youth & Society

and females. In the first step, violence at age 19 and 20 years was regressed on seventh-grade aggression, poverty status, and race. After adding adultrole statuses in the second step, we concluded that adult roles helped account for associations between early and later violence if coefficients for seventh-grade aggression became nonsignificant and substantially diminished in the second step of the model. We examined interactions between adult roles to address the current study’s second research question. We regressed young adult violence on all study variables and on two-way interactions between each adult role (for a total of three interaction terms in each model).

Results Bivariate Results As shown in Table 1, higher levels of seventh-grade aggression were positively associated with becoming a parent by age 19 and 20 years for males, χ2(1) = 23.7, p < .001; and females, χ2(1) = 15.3, p < .001. The proportions of males who became parents by age 19 and 20 years were 13%, 21%, and 35%, respectively, for those with low, moderate, and high levels of aggression in seventh grade. The proportions of females who were parents by age 19 and 20 years were 31%, 39%, and 58%, respectively, for females with low, moderate, and high levels of aggression in seventh grade. Moderate to high seventh-grade aggression was also associated with living independently at age 19 and 20 years among males, χ2(1) = 6.5, p < .05. Thirty-seven percent of males with low levels of aggressive behavior in seventh grade were living independently at age 19 and 20 years, compared to 49% of males with moderate or high levels of seventh-grade aggression. Adolescent aggressive behavior was also associated with work and schooling statuses at age 19 and 20 years for males, χ2(6) = 27.9, p < .001; and females, χ2(6) = 29.9, p < .001. Among females with high levels of aggressive behavior in seventh grade, 33% were behind in school or not in school and not working full-time, and 40% were behind in school or not in school and working full-time, which are substantially larger than the respective 18% and 31% of females with low levels of seventh-grade aggression. Among females with high levels of early adolescent aggressive behavior, 16% were seeking advanced education and not working full-time, and 11% were seeking advanced education and working full-time, compared to 29% and 22%, respectively, of those with low levels of aggression during early adolescence. Among males with high levels of seventh-grade aggression, Downloaded from http://yas.sagepub.com at JOHNS HOPKINS UNIV on October 3, 2008

245

Downloaded from http://yas.sagepub.com at JOHNS HOPKINS UNIV on October 3, 2008

None Any χ2 test (df)

Young adulthood violence

Seventh-grade aggression Low Moderate High χ2 test (df)

Males

78 (.20) 37 (.33) 8.1** (1)

21 (.13) 36 (.21) 58 (.35) 23.7*** (1)

Parent

161 (.42)x 62 (.55) 6.6* (1)***

60 (.37) 83 (.49) 80 (.49) 6.5* (1)***

Live Independently

95 (.25) 43 (.38)

29 (.18) 58 (.34) 51 (.31)

Not in or Behind in School, and not Working Full-Time

184 (.48)x 59 (.53)

75 (.47) 76 (.45) 92 (.55)

Not in or Behind in School, and Working Full-Time

59 (.15) 4 (.04) 19.6*** (3)

32 (.20) 18 (.11) 13 (.08) 27.9*** (6)

Advanced Education, and not Working Full-Time

Work and/or Schooling Status

Table 1 Summary of Bivariate Associations Between Adult Roles and Aggression in Seventh Grade and Violence at Age 19 and 20 Years (males: N = 497, females: N = 578)

(continued)

47 (.12) 6 (.05)

25 (.16) 18 (.11) 10 (.06)

Advanced Education, and Working Full-Time

246

Downloaded from http://yas.sagepub.com at JOHNS HOPKINS UNIV on October 3, 2008

201 (.32)x 32 (.36) 1.02 (1)

61 (.31). 74 (.39) 98 (.58) 15.3*** (1)

Note: Number (row percentage) shown. *p < .05. **p < .01. ***p < .001.

None Any χ2 test (df)

Young adulthood violence

Seventh-grade aggression Low Moderate High χ2 test (df)

Females

Parent

212 (.43)x 47 (.52) 2.37 (1)***

88 (.45) 82 (.43) 89 (.46) .4 (1)*

Live Independently

131 (.27)x 31 (.34)

35 (18)i 64 (.33) 63 (.33)

Not in or Behind in School, and not Working Full-Time

Table 1 (continued)

167 (.34)x 40 (.44)

61 (.31) 70 (.36) 76 (.40)

Not in or Behind in School, and Working Full-Time

108 (.22)x 12 (.13) 10.64* (3)

56 (.29) 34 (.17) 30 (.16) 29.9*** (6)

Advanced Education, and not Working Full-Time

Work and/or Schooling Status

82 (.17) 7 (.08)

42 (.22) 27 (.14) 20 (.11)

Advanced Education, and Working Full-Time

Roche et al. / Early Entries Into Adult Roles

247

31% were behind in school or not in school and not working full-time, and 55% were behind in school or not in school and working full-time, compared to 18% and 47%, respectively, of males with low levels of aggression in seventh grade. Among males with high levels of aggression in early adolescence, 8% were seeking advanced education and not working full-time, and 6% were seeking advanced education and working full-time, compared to 20% and 16%, respectively, of males with low seventh-grade aggression. Violence among young adult males and, to some extent, among young adult females, was associated with adult roles. One third of males who were parents reported violence in the past year at age 19 and 20 years; this compares to one fifth of males who were not parents, χ2 (1) = 8.1, p < .05. In addition, more than one half (55%) of males who lived independently at age 19 and 20 years reported having engaged in violence during the previous year, compared to 42% of males who did not live independently, χ2(1) = 6.6, p < .05. Work and schooling statuses at age 19 and 20 years were also associated with males’ violence, χ2(3) = 19.6, p < .001; and females’ violence, χ2(3) = 10.6, p < .05, during young adulthood. Among young adult females reporting any past-year violence, 34% were behind in school or not in school and not working full-time, and 44% were behind in school or not in school and working full-time, compared to 13% of those seeking advanced education and not working full-time and 8% of those seeking advanced education and working full-time. Among females not reporting any violence at age 19 and 20 years, these proportions were 27%, 34%, 22%, and 17%. Among males who reported violence at age 19 and 20 years, 38% were behind in school or not in school and not working full-time and 25% were behind in school or not in school and working full-time, compared to 4% of those getting advanced education and not working full-time and 5% of those getting advanced education and working full-time. Among males not reporting violence at age 19 and 20 years, these proportions were 25%, 48%, 15%, and 12%.

Multivariate Results Tables 2 and 3 present results from multivariate logistic regression analyses addressing the study’s first research question regarding whether or not adult roles account for associations between aggression during early adolescence and violence during young adulthood. Males whose seventhgrade aggression scores were moderate or high were 1.9 times more likely to report violence at age 19 and 20 years than males whose seventh-grade aggression scores fell within the lowest tertile (for moderate aggression, the 95% confidence interval [CI] was 1.1 to 3.3, p < .05; for high aggression,

Downloaded from http://yas.sagepub.com at JOHNS HOPKINS UNIV on October 3, 2008

248

Youth & Society

the 95% CI was 1.1 to 3.4, p < .01; Model 1, Table 3). When accounting for the acquisition of adult roles by age 19 and 20, years, associations between males’ seventh-grade aggression and violence in young adulthood substantially diminished in magnitude and were no longer statistically significant (Model 2, Table 3). Findings for adult roles indicated that males who were living independently at age 19 and 20 years were 1.6 times (95% CI: 1.0 to 2.5, p < .05) more likely to report having engaged in violence during the previous year. Males seeking advanced education and not working full-time were .16 (95% CI: .05 to .5, p < .01) times as likely as those not in school or behind in school and not working to engage in violence; and males seeking advanced education and working full-time were .34 (95% CI: .1 to .9, p < .05) times as likely as males not in school or behind in school and not working full-time to engage in violence. After accounting for adult roles, demographic attributes were significantly related to violence in young adulthood for males. White, non-Hispanic males were one half as likely (odds ratio [OR] = .52, 95% CI: .28 to .97, p < .05) as Black, non-Hispanic males to engage in violence, and males who were poor in seventh grade were .59 times as likely (95% CI: .35 to .98, p < .05) as males who were not poor to engage in violence at age 19 and 20 years. Findings for females indicated that the likelihood of any violence in young adulthood was higher among females who were more aggressive in seventh grade. Females with moderate and high levels of aggression were, respectively, 3.1 (95% CI: 1.6 to 6.1, p < .01) and 3.8 (95% CI: 1.9 to 7.5, p < .01) times more likely to report physical fighting in young adulthood than were females with low levels of aggression in seventh grade (Model 1, Table 3). The increased likelihood of aggression in young adulthood associated with females’ seventh-grade aggression scores remained virtually the same when accounting for females’ participation in adult roles. Among adult roles, only parenthood was related to females’ reports of violence at age 19 and 20 years. Females who were parents were 56% (95% CI: .34 to .93, p < .05) as likely as females who were not parents to report violence at the ages of 19 or 20 years. Although females who were poor in seventh grade were 1.8 (95% CI: .98 to 3.2, p < .05) times more likely than females who were not poor to report violence at age 19 and 20 years, the OR diminished in magnitude and was no longer significant when accounting for adult roles. We next ran models addressing our second research question regarding interactions among adult roles. Among males, there were no statistically significant two-way interactions between the three adult roles. Among females, two of the three interaction terms reached statistical significance (results are shown in Model 3 of Table 3 and are illustrated in Figures 1 and 2).

Downloaded from http://yas.sagepub.com at JOHNS HOPKINS UNIV on October 3, 2008

Roche et al. / Early Entries Into Adult Roles

249

Table 2 Summary of Multiple Logistic Regression Predicting Violence at Age 19 and 20 Years for Males (N = 497) Variable Seventh-grade aggressiona Moderate High White, non-Hispanicb Poorc Parent by age 19 and 20d Live independently at age 19 and 20f Not in or behind in school, working full-time Advanced education, not working full-time Advanced education, working full-time Model χ2 (df)

Odds Ratio

95% CI

Odds Ratio

95% CI

1.85* 1.93* .60 .84 — —

1.06, 3.25 1.11, 3.37 .33, 1.08 .53, 1.36 — —

1.52 1.44 .52* .59* 1.57 1.60*

.85, 2.72 .81, 2.56 .28, .97 .35, .98 .95, 2.59 1.01, 2.52

—

—

.76

.47, 1.25

—

—

.16**

.05, .48

—

—

.34*

.13, .87

12.2 (6)*

28.8 (5)***

Note: Reference groups are the following: a. Seventh-grade aggression: Low; b. African American; c. Not poor; d. Not a parent; e. Do not live independently; f. Work and/or schooling: Not in school/behind in school and do not work full-time. *p < .05. **p < .01. ***p < .001.

As illustrated in Figure 1, the odds that females would report violent behavior at age 19 and 20 years was significantly higher among females who were living independently and not in school or behind in school and working full-time. When compared to females who were still living with family, not in school or behind in school and not working full-time, females who were living independently and not in school or behind in school and working full-time were 3.9 times more likely to report having engaged in violence within the previous year. Independent living was not associated with increased violence for females in any other work and schooling categories. The current study’s second significant interaction effect for females is illustrated in Figure 2. We retained this interaction term in our model, yet the coefficient just failed a test of statistical significance at the probability level equal to .05 (as opposed to p < .05). Overall, advanced education was associated with less violence, regardless of females’ work. However, among females not in school or behind in school and working full-time, those who were parents reported

Downloaded from http://yas.sagepub.com at JOHNS HOPKINS UNIV on October 3, 2008

250

Downloaded from http://yas.sagepub.com at JOHNS HOPKINS UNIV on October 3, 2008

Seventh grade aggressiona Moderate High White, Non-Hispanicb Poorc Parent by age 19 and 20d Live independently at age 19 and 20e Work and or schooling at age 19 and 20f Not in or behind in school, working full-time Advanced education, not working full-time Advanced education, working full-time

Variable

3.09*** 3.76*** 1.34*** 1.81***

Odds Ratio

2.80*** 3.66*** 1.38*** 1.78*** .56** 1.40***

1.10***

.58**

.47**

—

—

—

Odds Ratio

1.56, 6.13 1.89, 7.49 .79, 3.06 1.06, 3.33 — —

95% CI

.19, 1.11

.28, 1.25

.65, 1.89

1.40, 5.60 1.81, 7.41 .69, 2.75 .98, 3.20 .34, .93 .88, 2.31

95% CI

Table 3 Summary of Multiple Logistic Regression Predicting Violence at Age 19 and 20 Years for Females (N = 578)

.57**

.59**

.88**

2.84*** 3.67*** 1.25*** 1.72*** .91** .81**

Odds Ratio

.16, 1.98

.20, 1.73

.35, 2.19

1.40, 575 1.80, 7.48 .62, 2.53 .95, 3.13 .40, 2.06 .36, 1.84

95% CI

251

Downloaded from http://yas.sagepub.com at JOHNS HOPKINS UNIV on October 3, 2008

—

—

—

—

—

—

—

—

40.9 (5)*

—

—

28.9 (6)***

—

—

51.2 (9)***

.75**

1.15***

3.51***

.88**

.35†*

.35**

.11, 5.02

.26, 5.11

1.14, 10.8

.13, 6.01

.12, 1.09

.12, 1.09

Note: Reference groups are the following: a. Seventh-grade aggression: Low; b. African American; c. Not poor; d. Not a parent; e. Do not live independently, f. Work and/or schooling: Not in school or behind in school and do not work full-time. †p = .05. *p < .05. **p < .01. ***p < .001.

Model χ2 (df)

Two-way interaction effects Parent by age 19 and 20 years & Not in or behind in school, work full-time Advanced education, not working full-time Advanced education, working full-time Independent living by age 19 and 20 years & Not in or behind in school, working full-time Advanced education, not working full-time Advanced education, working full-time

252

Youth & Society

Figure 1 Odds of Violent Behavior Among Females by Work and/or Schooling Status and Independent Living at Age 19 and 20 Years 4.5 4 3.5 3 2.5 2 1.5 1 0.5 0 Not Living Independently

Living Independently

Behind in School and Not Working Full Time Behind in School and Working Full Time Advanced Education and Not Working Full Time Advanced Education and Working Full Time

less violence within the previous year. Among this group, females who had become parents by age 19 and 20 years were 29% as likely to report pastyear violence as their counterparts who were not parents.

Discussion As in other studies examining life-course trajectories of aggressive and antisocial behaviors, youth in the current study who were more aggressive during early adolescence were at increased risk of engaging in violence during young adulthood. Our aim here was to examine whether or not early entries into adult roles helped explain the processes by which youth living in poor, urban neighborhoods continue or reduce involvement in aggression

Downloaded from http://yas.sagepub.com at JOHNS HOPKINS UNIV on October 3, 2008

Roche et al. / Early Entries Into Adult Roles

253

Figure 2 Odds of Violent Behavior Among Females by Work and/or Schooling Status and Parenthood at Age 19 and 20 Years 1.2

1

0.8

0.6

0.4

0.2

0 Not a Parent

Parent

Behind in School and Not Working Full Time Behind in School and Working Full Time Advanced Education and Not Working Full Time Advanced Education and Working Full Time

later in the life course. Our rationale for focusing on adult roles was twofold. First, our previous study of youth included in the current study provided compelling evidence of a link between independent roles (e.g., preparing meals, babysitting, doing chores, working, unsupervised time with peers) and changes in fighting and weapon use during early adolescence (Roche et al., 2003). Second, life-span studies supported the idea that adult roles are related to deviant behaviors and that early entries into adult roles are salient to the lives of youth in poor urban areas (Burton et al., 1995; Nielsen, 1999). Although we included males and females in the current study, we used gender-specific measures of aggression and violence because of the small proportion of females reporting serious violent behavior (e.g., gun carrying, using force to steal) in young adulthood. We may have benefited from a measure tapping into relational forms of aggressive behavior as previous

Downloaded from http://yas.sagepub.com at JOHNS HOPKINS UNIV on October 3, 2008

254

Youth & Society

research suggests that this better captures aggression among females than indicators of physical fighting and violence (Crick & Grotpeter, 1995). We were also unable to gain a meaningful understanding of associations between aggression and marriage because so few youth in the current sample reported ever having been married. The low prevalence of marriage among this sample, coupled with previous research showing fewer benefits of marriage for African Americans compared to Whites and Latinos, raises questions about the relevance of marriage as an adult role for some African Americans, at least in the very early years of adulthood. Despite these limitations, we did find support for the idea that adult roles are relevant to differences in early adolescent aggression among males and females. For example, males and females who were rated by teachers as being more aggressive in seventh grade were more likely than those who were less aggressive to have children by age 19 and 20 years. In addition, males who were more aggressive were more likely than males who were less aggressive to live independently at age 19 and 20 years. Contrary to our expectations, early aggressive behavior was not associated with entry into full-time work. Rather, aggression was positively associated with an individual’s schooling status in young adulthood. Youth who were more aggressive were less likely to be seeking advanced education at age 19 and 20 years, regardless of whether or not they worked full-time, and they were more likely to be behind in school or not in school, regardless of whether or not they worked full-time. The greater salience of aggression to education, as compared to employment, is consistent with findings from previous research showing that schooling, but not work, was related to adult criminal behavior (Horney et al., 1995). Perhaps full-time work is an insufficient indicator of a precocious adult role among those living in economically distressed areas because youth in these contexts have unusually great financial needs. Information about the reasons why youth in the current sample worked full-time and what their plans were for future jobs and education may have enabled us to develop a measure of work that more accurately captured a precocious entry into the labor force. Regardless of why and how youth who are more aggressive entered adult roles precociously, each was relevant to violence in young adulthood. Males who had high levels of aggression during early adolescence, for example, were 2.2 times more likely than those with low levels of aggression to engage in violence in young adulthood. The risk of serious violence associated with a high level of early adolescent aggression declined 60% when accounting for males’ participation in adult roles, namely greater tendency

Downloaded from http://yas.sagepub.com at JOHNS HOPKINS UNIV on October 3, 2008

Roche et al. / Early Entries Into Adult Roles

255

of males who are aggressive to live independently and to not seek advanced education. Young men who lived independently at age 19 and 20 years were 60% more likely than those living with family to engage in serious violence. And, when compared to men who were neither seeking advanced education nor working full-time, young adult males seeking advanced education who were not working full-time were 16% as likely to engage in serious violence, and males seeking advanced education and working fulltime were 34% as likely as to engage in serious violence. Parenthood was unrelated to males’ reports of serious violence at age 19 and 20 years. The current study does not allow us to draw causal inferences between early involvement in aggression and later independent living arrangements and reduced schooling. Other family or neighborhood characteristics may be related to both. Families, for example, may not allow males who are aggressive to continue to live with them. Other studies have shown that youth who are aggressive or deviant are more likely to do poorly in school and drop out (Ensminger & Slusarick, 1992; Smith et al., 2000). Similarly, we do not know from our findings why independent living and being behind in school or not in school are associated with acts of violence. With these caveats in mind, there is certainly convincing theoretical support for the possibility of causal pathways from early adolescent aggressive behavior to independent living and reduced schooling and, in turn, to serious violence among males in young adulthood. Boys whose teachers perceive them as engaged in high levels of fighting and disruptive behaviors are likely to, consequently, experience problems in schooling leading to dropout or grade retention. Boys who are more aggressive during early adolescence are also more likely to develop stronger affiliations with deviant peer groups rather than with family (Patterson & Dishion, 1985) that, in turn, would inspire them to live independently rather than with family at age 19 and 20 years. As suggested by previous theory and research, living independently and a lower attachment to school provide young men with fewer informal social controls and bonds to social institutions that, in turn, act as deterrents to one’s decision to carry a gun or use force to steal something (Sampson & Laub, 1990). Although parenthood might also deter males’ engagement in violence for these same reasons, it is likely that a relatively small proportion of men in the current sample actually lived with their children. Thus, parenthood would not inspire males to act responsibly or as positive role models. Early adolescent aggressive behavior was associated with females’ increased risk of violence in young adulthood. Females with moderate and high levels of aggressive behavior in seventh grade were, respectively, 3.1 and

Downloaded from http://yas.sagepub.com at JOHNS HOPKINS UNIV on October 3, 2008

256

Youth & Society

3.8 times more likely than those with low levels of aggressive behavior to report violence in young adulthood. Precocious entries into adult roles did not, however, alter the odds that early adolescent aggression would be associated with later violence. Among adult roles, only parenthood was associated with females’ reports of violence during young adulthood. Consistent with our expectation that a family-oriented adult role such as parenthood would be associated with declines in violence, females who had become parents by age 19 and 20 years were 56% as likely as other females to have been involved in violence during young adulthood. Thus, girls who were highly aggressive during early adolescence were more likely than others to become parents at a young age; yet, young mothers did not experience declines in violence as a result. Other studies have raised provocative questions about the extent to which early childbearing leads to negative consequences among African American youth living in this country’s most economically disadvantaged, urban neighborhoods. Geronimus (1991) argued that delayed childbearing enables White non-Hispanic, but not African Americans, living in high-poverty neighborhoods to gain educational and employment advantages. She contended that African American females who are low income who delay childbearing face the same limited opportunities for jobs, marriage, and education as their counterparts who bear children during adolescence, but that they have fewer extended kin to rely on for support and assistance in raising their children because the kin are now often experiencing their own health problems or are no longer alive (Geronimus). Other studies of African American populations, though, have found that early childbearing is not advantageous to education, family formation, or employment (Jacobsen, 1998; Kellam et al., 1982). Although our findings suggest a protective effect of early childbearing against increased risks of violence in young adulthood, these findings must be evaluated in the context of a young woman’s work and educational status. This brings us to findings addressing the current study’s second research question regarding interactions among adult roles. We found that working mothers with low levels of education engaged in less violence than other women with low levels of education. Having children did not distinguish violence risks among women seeking advanced educational opportunities, whose overall risks were lower when compared to women with less education. An obvious explanation for why mothers who work full time without advanced education were buffered from the increased risks of violence faced by other women with less education is that they simply did not have time for involvement in activities other than working and caring for their children. This supports the idea that increased responsibility may mitigate the likelihood of aggressive behavior. Given that many neighborhoods in

Downloaded from http://yas.sagepub.com at JOHNS HOPKINS UNIV on October 3, 2008

Roche et al. / Early Entries Into Adult Roles

257

which study participants lived had high levels of crime and unemployment, women who were not parents and full-time workers likely had more time to hang out with others who were unemployed and not in school. As a result, they may have experienced more opportunities for becoming involved in delinquent behaviors such as fighting. Full-time work also modified the association between independent living and fighting among young adult females with low levels of education. When not seeking advanced education, young women who worked full-time and lived independently experienced substantial increases in the risk of violence. These women were four times more likely to have been involved in violence at age 19 and 20 years than females who lived independently and did not work full-time (and were not seeking advanced education), who did not live independently, and who were seeking advanced education. In interpreting these findings, it is useful to consider that the jobs held by urban women who were low income without advanced education commonly offer low wages and few benefits and provide little stability in the job market. Women working in these positions often, therefore, experience considerable psychological and financial stress in their lives. Living with family, as opposed to independently, may therefore buffer women against the stresses of low-wage work and the ensuing risks of becoming involved in violence by providing women with greater emotional and financial support from families. Women who are seeking advanced education may not experience the increased risks posed by independent living or full-time work because of the contextual advantage of attending a college or university, as compared to high school or no school. The fact that early parenthood and independent living, taken alone, were not associated with increased risks among females demonstrates the importance of evaluating early adult roles within a larger context of individuals’ lives. Clearly, for females, one’s educational status and participation in the labor force have important bearings on the extent to which parenthood and independent living pose increased risks of fighting at age 19 and 20 years. The interactive nature of adult roles is consistent with our previous findings demonstrating that paid work modified influences of familial and peer independent roles on changes in young adolescent’s aggressive behavior (Roche et al., 2003). There are important limitations of the current study. First, we know little about the timing and dynamic changes in adult role acquisition among respondents. As mentioned in our discussion of findings for males, adult roles such as independent living and work and/or schooling may have been the consequence rather than cause of violence and aggression at age 19 and 20 years. Information about changes in work and schooling over time would

Downloaded from http://yas.sagepub.com at JOHNS HOPKINS UNIV on October 3, 2008

258

Youth & Society

also have provided us with a better understanding of whether or not individuals who were not in school or behind in school and working full-time had begun to work as a way to establish fiscal responsibility or were simply having to work while finishing school with the intention of continuing their education after completion of high school. Second, our findings for parenthood are limited because we do not know if the participant lived with the child. Parents who live with their children experience entirely different structures and responsibilities on a daily basis when compared to parents who do not live with, or have contact with, their children. Finally, the current study is limited with respect to sample attrition as 46% of original participants did not participate at age 19 and 20 years. Because those lost to attrition were more likely to be White, non-Hispanic as compared to African American, we likely had a lower prevalence of marriage and independent living than would be expected without this sample attrition (Aquilino, 1991; Buck & Scott, 1993). We expect that the current study findings represent a conservative estimate of how adult roles are associated with aggression and serious violence in young adulthood because the sample attrition was greater among youth who were more aggressive in first grade. In sum, the current study provides strong support for the relevance of adult role acquisition to aggressive and violent behaviors during early adolescence and the transition to adulthood among young people growing up in low-income, urban neighborhoods. Future studies should consider how adult roles relate to other domains of youth health and well-being within high-risk settings.

References Aquilino, W. S. (1991). Family structure and home-leaving: A further specification of the relationship. Journal of Marriage and the Family, 53, 999-1010. Aquilino, W. S. (1996). The life course of children born to unmarried mothers: Childhood living arrangements and young adult outcomes. Journal of Marriage and the Family, 58, 293-310. Arnett, J. J. (2001). Adolescence and emerging adulthood: A cultural approach. Upper Saddle River, NJ: Prentice Hall. Bachman, J. G., O’Malley, P. M., & Johnston, L. D. (1984). Drug use among young adults: The impacts of role status and social environments. Journal of Personality and Social Psychology, 47, 629-645. Broidy, L. M., Nagin, D. S., Tremblay, R. E., Bates, J. E., Brame, B., Dodge, K. A., et al. (2003). Developmental trajectories of childhood disruptive behaviors and adolescent delinquency: A six-site, cross-national study. Developmental Psychology, 39, 222-245. Buck, N., & Scott, J. (1993). She’s leaving home: But why? An analysis of young people leaving the parental home. Journal of Marriage and the Family, 55, 863-874.

Downloaded from http://yas.sagepub.com at JOHNS HOPKINS UNIV on October 3, 2008

Roche et al. / Early Entries Into Adult Roles

259

Burton, L. M., Allison, K. W., & Obeidallah, D. (1995). Social context and adolescence: Perspectives on development among inner-city African-American teens. In L. J. Crockett & A. C. Crouter (Eds.), Pathways through adolescence: Individual development in relation to social contexts (pp. 119-138). Hillsdale, NJ: Lawrence Erlbaum. Card, J. J., & Wise, L. L. (1978). Teenage mothers and teenage fathers: The impact of early childbearing on the parent’s personal and professional lives. Family Planning Perspectives, 10, 199-205. Centers for Disease Control and Prevention. (2002). Youth risk behavior trends from CDC’s 1991, 1993, 1995, 1997, and 1999 Youth Risk Behavior surveys. Retrieved August 1, 2002, from www.cdc.gov/nccdphp/dssh/yrbs/trned.htm Chen, Z., & Kaplan, H. B. (1999). Explaining the impact of family structure during adolescence on adult educational attainment: A longitudinal study. Applied Behavioral Science Review, 7, 23-40. Cicchetti, D., & Schneider-Rosen, K. (1984). Toward a transactional model of childhood depression. In D. Cicchetti & K. Schneider-Rosen (Eds.), Childhood depression: A developmental perspective (pp. 5-28). San Francisco: Jossey-Bass. Cooney, T. M., & Mortimer, J. T. (1999). Family structure differences in the timing of leaving home: Exploring mediating factors. Journal of Research on Adolescence, 9, 367-394. Crick, N. R., & Grotpeter, J. K. (1995). Relational aggression, gender, and social-psychological adjustment. Child Development, 66, 710-722. Dolan, L. J., Kellam, S. G., Brown, C. H., Werthamer-Larsson, L., Rebok, G. W., Mayer, L. S., et al. (1993). The short-term impact of two classroom-based preventive interventions on aggressive and shy behaviors and poor achievement. Journal of Applied Developmental Psychology, 14, 317-345. Ensminger, M. E., & Slusarick, A. (1992). Paths to high school graduation or dropout: A longitudinal study of a first grade cohort. Sociology of Education, 65, 95-113. Geronimus, A. T. (1991). Teenage childbearing and social reproductive disadvantage: The evolution of complex questions and the demise of simple answers. Family Relations, 40, 463-471. Gove, W. R. (1985). The effect of age and gender on deviant behavior: A biopsychological perspective. In A. S. Rossi (Ed.), Gender and the life course (pp. 115-144). New York: Aldine. Harris, A. R. (1977). Sex and theories of deviance: Toward A functional theory of deviant typescripts. American Sociological Review, 42, 3-16. Hirschi, T. (1969). Causes of delinquency. Berkeley: University of California Press. Hirschl, T. A., Altobelli, J., & Rank, M. R. (2003). Does marriage increase the odds of affluence? Exploring the life course probabilities. Journal of Marriage and the Family, 65, 927-938. Horney, J., Osgood, D. W., & Marshall, I. H. (1995). Criminal careers in the short-term: Intraindividual variability in crime and its relation to local life circumstances. American Sociological Review, 60, 655-673. Jacobsen, J. (1998). Does timing of first birth affect young adult outcomes in a population of at risk African American women? Unpublished doctoral dissertation, Johns Hopkins University, Baltimore. Kellam, S. G. (1990). Developmental epidemiologic framework for family research on depression and aggression. In G. R. Patterson (Ed.), Depression and aggression in family interaction (pp. 11-48). Hillsdale, NJ: Lawrence Erlbaum. Kellam, S. G., Adams, R. G., Brown, C. H., & Ensminger, M. E. (1982). The long-term evolution of the family structure of teenage and older mothers. Journal of Marriage and the Family, 44, 539-554.

Downloaded from http://yas.sagepub.com at JOHNS HOPKINS UNIV on October 3, 2008

260

Youth & Society

Kellam, S. G., Branch, J. D., Agrawal, K. C., & Ensminger, M. E. (1975). Mental health and going to school: The Woodlawn program of assessment, early intervention, and evaluation. Chicago: University of Chicago Press. Kellam, S. G., & Ensminger, M. E. (1980). Theory and method in child psychiatric epidemiology. In F. J. Earls (Ed.), Studies of children (pp. 145-180). New York: Prodist. Kellam, S. G., & Rebok, G. W. (1992). Building developmental and etiological theory through epidemiologically based preventive intervention trials. In J. McCord & R. E. Tremblay (Eds.), Preventing antisocial behavior: Interventions from birth through adolescence (pp. 162-195). New York: Guilford. Kellam, S. G., Werthamer-Larsson, L., Dolan, L. J., & Brown, C. H. (1991). Developmental epidemiologically based preventive trials: Baseline modeling of early target behaviors and depressive symptoms. American Journal of Community Psychology, 19, 563-584. Loeber, R., & Hay, D. (1997). Key issues in the development of aggression and violence from childhood to early adulthood. Annual Review of Psychology, 48, 371-410. Marini, M. M. (1985). Determinants of the timing of adult role entry. Social Science Research, 14, 309-350. Maugham, B., & Rutter, M. (1998). Continuities and discontinuities in antisocial behavior from childhood to adult life. In T. H. Ollendick & R. J. Prinz (Eds.), Advances in clinical child psychology (Vol. 20, pp. 1-47). New York: Plenum. Michael, R. T., & Tuma, N. B. (1985). Entry into marriage and parenthood by young men and women: The influence of family background. Demography, 22, 515-544. Moffitt, T. (1993). Adolescence-limited and life-cycle persistent antisocial behavior: A developmental taxonomy. Psychological Review, 100, 674-701. Newcomb, M., & Bentler, P. M. (1988). Consequences of adolescent dug use: Impact on the lives of young adults. Newbury Park, CA: Sage. Nielsen, A. L. (1999). Testing Sampson and Laub’s life course theory: Age, race/ethnicity, and drunkenness. Deviant Behavior: An Interdisciplinary Journal, 20, 129-151. Patterson, G. R., & Dishion, T. J. (1985). Contributions of families and peers to delinquency. Criminology, 23, 63-79. Patterson, G. R., Reid, J. B., & Dishion, T. J. (1992). Antisocial boys: A social interactional approach (Vol. 4). Eugene, OR: Castalia Publishing. Petras, H., Chilcoat, H. D., Leaf, P. J., Ialongo, N. S., & Kellam, S. G. (2004). Utility of TOCA-R scores during the elementary school years in identifying later violence among adolescent males. Journal of the American Academy of Child and Adolescent Psychiatry, 43, 88-96. Roche, K. M., Ensminger, M. E., Chilcoat, H., & Storr, C. (2003). Establishing independence in low-income urban areas: The relationship to adolescent aggressive behavior. Journal of Marriage and Family, 65(3), 668-680. Sampson, R., & Laub, J. (1990). Crime and deviance over the life course: The salience of adult social bonds. American Sociological Review, 55, 609-627. Smith, C. A., Krohn, M. D., Lizotte, A. J., McCluskey, C. P., Stouthamer-Loeber, M., & Weiher, A. (2000). The effect of early delinquency and substance use on precocious transitions to adulthood among adolescent males. Families, Crime and Criminal Justice, 2, 233-253. Tang, S. (1997). The timing of home leaving: A comparison of early, on-time, and late home leavers. Journal of Youth and Adolescence, 26, 13-23. Thompson, M. S., & Ensminger, M. E. (1989). Psychological well-being among mothers with school age children: Evolving family structures. Social Forces, 67, 715-730.

Downloaded from http://yas.sagepub.com at JOHNS HOPKINS UNIV on October 3, 2008

Roche et al. / Early Entries Into Adult Roles

261

Uggen, C. (2000). Work as a turning point in the life course of criminals: A duration model of age, employment, and recidivism. American Sociological Review, 67, 529-546. Werthamer-Larsson, L., Kellam, S. G., & Wheeler, L. (1991). Effect of first-grade classroom environment on child shy behavior, aggressive behavior, and concentration problems. American Journal of Community Psychology, 19, 585-602. Windle, M. (1990). A longitudinal study of antisocial behaviors in early adolescence as predictors of late adolescent substance use: Gender and ethnic group differences. Journal of Abnormal Psychology, 99, 86-91. Wu, L. L., & Martinson, B. C. (1993). Family structure and the risk of a premarital birth. American Sociological Review, 58, 210-232.

Kathleen M. Roche is an assistant scientist in the Department of Population and Family Health Sciences at the Johns Hopkins Bloomberg School of Public Health. Her research has focused on understanding social contextual influences on adolescents growing up in lowincome families and high-risk communities. Margaret E. Ensminger is a professor in the Department of Health, Behavior, and Society at the Johns Hopkins Bloomberg School of Public Health. She has focused her research on understanding pathways to social adaptation and psychological well-being across the life course in low-income urban communities. Nicholas Ialongo is an associate professor in the Department of Mental Health at the Johns Hopkins Bloomberg School of Public Health. He is a child and family clinical psychologist by training, and his research interests include the prevention of and early intervention with mental disorders and substance abuse in children and adolescents. Jeanne M. Poduska is a principal research scientist and the deputy director of the Center for Integrating Education and Prevention Research in Schools at the American Institutes for Research. Her research focuses on developmental epidemiology, randomized field trials in prevention, services research, and the integration across these areas. Sheppard G. Kellam is a senior research fellow at the American Institutes for Research, and director of the A.I.R. Center for Integrating Education and Prevention Research in Schools. He is Professor Emeritus Johns Hopkins Bloomberg School of Public Health. His research is about concepts and methods in prevention science, developmental epidemiology, and developing and testing preventive interventions in randomized field trials

Downloaded from http://yas.sagepub.com at JOHNS HOPKINS UNIV on October 3, 2008