Parenting Practices and Peer Group Affiliation in Adolescence

B. Bradford Brown University of Wisconsin—Madison

Mna Mounts University of Illinois at Urbana-Champaign Susie D. Lamborn University of West Florida

Laurence Steinberg Temple University

BROWN, B. BRADFORD; MOUNTS, NINA; LAMBORN, SUSIE D.; and STEINBERG, LAURENCE. Parenting Practices and Peer Group Affiliation in Adolescence. CHILD DEVELOPMENT, 1993, 64, 467-482. Social scientists have often assumed that parental influence is sharply curtailed at adolescence because of the rising counterinfluence of peer groups, over which parents have little control. The present study tested a conceptual model that challenged this view by arguing that parents retain a notable but indirect infiuence over their teenage child's peer associates. Data from a sample of 3,781 high school students (ages 15—19) indicated that specific parenting practices (monitoring, encouragement of achievement, joint decision making) were significantly associated with specific adolescent behaviors (academic achievement, drug use, self-reliance), which in turn were significantly related to membership in common adolescent crowds (jocks, druggies, etc.). Findings encourage investigators to assess more carefully parents' role in adolescents' peer group affiliations.

Among the many transformations in across early adolescence, susceptibility to family relationships expected to occur as a peer pressure increases while reliance upon cbild enters adolescence, one of the most parents' opinions and advice seems to de- worrisome for adults is the shift in reference cline (Berndt, 1979), as well as by research group orientation from parents to the peer indicating that association with deviant group (Gecas & Seff, 1990). Parents often ex- peers is one of the strongest predictors of press concern that their adolescent child adolescent deviant activity (Elliott, Hui- will "fall into the wrong crowd" and be per- singa, & Ageton, 1985). suaded by peer pressure to engage in behav- iors that are self-destructive and/or counter There is, however, evidence that par- to parental expectations. Such concern may ents do retain a substantial measure of in- seem justified by studies indicating that, fluence over the attitudes and activities of

A version of this paper was presented at the biennial meeting of the Society for Research in Ghild Development, Seattle, 1991. The study on which this report is based was supported by grants to B. Bradford Brown from the Spencer Foundation through the Wisconsin Genter for Education Research, University of Wisconsin—Madison; to B. Bradford Brown and Laurence Steinberg from the U.S. Department of Education through the National Center on Effective Secondary Schools, Wisconsin Center for Educational Research, University of Wisconsin— Madison; and to Sanford M. Dombusch. and P. Herbert Leiderman from the Spencer Foundation through the Genter for the Study of Families, Youth, and Ghildren at Stanford University. The opinions expressed in this paper are those of the authors and do not necessarily reflect the opinions of the funding agencies. The authors gratefully acknowledge the helpful comments of three anonymous reviewers on earlier versions of this paper, as well as the assistance of Bih-Hui Huang in data analyses and Lori Folgert in preparing the manuscript. Address correspondence to the first author at the Department of Educational Psychology, University of Wisconsin— Madison, 1025 West Johnson Street, Madison, WI 53706. [Child Development, 1993, 64,467-482. © 1993 by the Society for Research in Child Development, Inc. All rights reserved. 0009-3920/93/6402-0008$01.00] 468 Child Development their teenage offspring. We believe this evi- Second, in contrast to the mid-century dence is key to demonstrating the significant image of a unified peer culture, ethnogra- role that parents play in their child's choice phers have consistently discovered a diverse of peer group associates during adolescence. array of peer groups in American high The present study tests a conceptual model schools (Buff, 1970; Eckert, 1989; Eder, in which parenting behaviors affect adoles- 1985; Larkin, 1979; Schwendinger & cents' peer group affiliations significantly Schwendinger, 1985). An elite group similar but indirectly, through their infiuence on ad- to Coleman's (1961) "leading crowd" is al- olescent behavior patterns. ways found, although it is often divided into two somewhat different crowds. "Populars" The Traditional View and Contemporary are portrayed as socially competent individ- Counter evidence uals with a strong commitment to academic At mid-century, social scientists pro- achievement but also moderate involvement moted a perspective that emphasized the in delinquent behavior and illicit drug use. competitive or combative role that parents "Jocks" are quite similar but less academi- and peers play in the lives of teenagers. Par- cally oriented and more focused in their sons (1942) portrayed the adolescent social drug use on alcohol, which they sometimes system as a monolithic youth culture whose use to excess. Counterposed against these value system is clearly at odds with adult crowds is a more alienated group—"drug- expectations. Riesman (1961) contended that gies," "burnouts," "greasers," etc.—that is parents had abdicated much of their author- not only heavily involved in drug use and ity over adolescent offspring to the school deviant activities but also inattentive to and the peer group. Thus, it was common schoolwork and often hostile toward school to portray parents and peers as competitors adults; yet, group members seem to main- for teenagers' allegiance, with parents tain a fairly strong self-image. Balanced be- relatively powerless—in many domains, at tween these groups are the "normals," aver- least—to countermand the infiuence of peer age, or "in-between" students who seem to group pressures (Bowerman & Kinch, 1959; avoid deviant activities but otherwise are Brittain, 1963; Coleman, 1961; Floyd & not clearly distinctive on any particular trait. South, 1972). In most studies there is also a group of high More recent studies, however, have pro- achievers, the "brains" or "eggheads" or vided a very different image of parent and "intellectuals," who thrive on academics, peer group infiuences on adolescents. In the forge close relationships with school adults, first place, investigators have found that pa- and studiously avoid drugs and deviant ac- rental influence on children's behavior re- tivities. Their self-confidence is bolstered by mains extensive in adolescence. Through academic achievements but also eroded by particular child-rearing practices, parents their marginal standing in the peer status can have a substantial impact on adolescent system. Most schools also feature a socially behaviors of major concern to adults, such as inept crowd—"loners" or "nerds"—whose school achievement patterns, drug use and members are generally low in socid status deviance, and self-concept. For example, and, consequently, self-esteem. Their aca- outside of the child's intellectual ability, pa- demic achievement levels are variable, but rental expectation for achievement is one of they seem to shy away from deviant activ- the strongest predictors of adolescents' aca- ities. demic achievement levels (Featherman, 1980; Seginer, 1983). Parental monitoring The ethnographic studies not only em- has a particularly powerful infiuence on ado- phasize the diversity of life-styles that ado- lescent delinquency and drug use (Coombs lescents pursue with peers, they also empha- & Landsverk, 1988; Loeber & Dishion, size the limits of peers' ability to direct 1983) as well as on academic achievement teenagers' behavior. Adolescents do not hap- (Dombusch, Ritter, Leiderman, Roberts, & hazardly fall into one crowd or another and Fraleigh, 1987). Parental efforts to engage then fall victim to the normative pressures their adolescent in joint decision making of that crowd. Instead, they are "selected seem to foster self-confidence and self- into" a particular crowd by virtue of the rep- reliance (Baumrind, 1991b; Grotevant & utation they establish among peers (Brown, Cooper, 1986; Hauser et al., 1984), enhance 1990). One does not become a jock by quirk academic performance (Dombusch et al., of fate or even simply by honing one's ath- 1987; Yee & Flanagan, 1985), and discour- letic skill and making the basketball team. age excessive or abusive use of drugs One must be perceived by peers as acting (Glynn, 1984) or delinquent activities (Dom- primarily in "jock-like" ways: being inter- busch & Ritter, 1991). ested in sports, getting adequate but not out- Brown et al. 469 standing grades, exuding self-confidence but volved in drug use (Kandel & Andrews, remaining compliant with adult authority, 1987; Oetting & Beauvais, 1987) or delin- shying away from drugs except for alcohol, quency (Marcos, Bahr, & Johnson, 1986; and so on. There is, to be sure, pressure to Massey & Krohn, 1986). Others have sug- conform to crowd norms (Clasen & Brown, gested a different causal arrangement of 1985) and to select friends from fellow these variables: parental detachment from crowd members (Eckert, 1989; Eder, 1985), early adolescents promotes antisocial behav- but these do not direct adolescents to new iors, and youngsters with such behaviors behaviors as much as they reinforce existing seem to band together in antisocial cliques dispositions—dispositions that helped di- (Cairns, Cairns, Neckerman, Gest, & Ga- rect the adolescent to a particular crowd in riepy, 1988; Dishion, 1990). To date, how- tlie first place. ever, such studies have been largely con- fined to deviant behavior and deviantly In sum, parents retain substantial influ- oriented peer groups. ence over their child's behavior in ado- lescence, and peer influences are not pre- The intent of this study was to expand dominantly antisocial but quite variable, consideration of parental influences on ado- contingent on the norms of the peer crowd lescent peer group affiliation to a broader to which the adolescent belongs. Further, array of peer groups and to consider both crowd affiliation is not a matter of chance antisocial and prosocial behaviors. The as- association but a function of the reputation sumption that parenting behaviors have a di- one establishes among peers by virtue of rect effect on teenagers' peer relationships one's background and behavior. follows neatly from similar research on younger age groups (e.g., Parke & Bhavna- Linking Parenting Practices gri, 1989). Nevertheless, the evidence re- to Peer Group Affiliations viewed above, indicating that parents infiu- Ethnographers have provided a detailed ence many behaviors by which adolescents analysis of the characteristics of adolescent are assigned to crowds, seems to favor a con- crowds but have been remarkably inatten- ceptual model in which parents play a sig- tive to the family characteristics of crowd nificant but indirect role in adolescent peer members. By the same token, researchers group affiliations (see Fig. 1). According to who have detailed associations between par- this model, specific parenting behaviors are enting strategies and adolescent behaviors significantly associated with specific adoles- usually have not concerned themselves with cent characteristics, which in turn predict adolescents' peer group relationships. This the peer group widi which the adolescent fosters the impression that the family and is associated. Peer group pressures serve to peer group are quite separate worlds for reinforce behavior patterns by which adoles- teenagers, perhaps because by this age par- cents come to be associated with a particular ents have lost their ability to affect peer asso- crowd. By fostering certain traits in their ciations. Recent studies of adolescent devi- children, parents essentially direct a child ance counter this impression, but they also toward a particular peer group, and thus ex- raise questions about the way in which par- ercise some control over the type! of peer ents influence peer group affiliations. Some group infiuences to which their child is ex- investigators have found that if parents posed. model deviant behavior or fail to maintain close relationships with their teenager, the To test the conceptual model, we fo- child is more likely to drift into deviant peer cused on three parenting practices— groups and, as a consequence, be more in- emphasis on academic achievement, moni-

Parenting Adoiescent's Crowd practices behaviors affiiiation i Crowd norms& peer pressure

FIG. 1.—Conceptual model of the connections between parenting behaviors, peer group influ- ences, and adolescent behavior. 470 Child Development toring, and efforts to engage the child in joint tion) but also included measures of family decision making—that have been shown in relationships and parenting behaviors, peer previous studies to be significant parental relationships, deviant activities, and psycho- influences on adolescent behavior. We also logical well-being. Because of its length, the selected three adolescent behaviors— questionnaire was divided into two sections, academic achievement, drug use, and self- which were administered on separate days. reliance—that are not only highly salient to Students also received peer ratings of crowd researchers, educators, and parents, but also affiliation, based on the Social Type Rating instrumental in defining adolescent peer (STR) procedure (Brown, 1989) that is ex- groups. We relied upon adolescents' own plained in the next section. STR ratings descriptions of their social system to define were obtained for all students in the Mid- the crowds to be studied. Because the western schools but, because of time con- study's basic intent was to examine the man- straints, only freshmen and sophomores in ner in which parents influence peer group the West Coast schools. affiliation, the portion of the conceptual model specifying peer group influences on Analyses in the present study were con- adolescent behavior patterns (the "feedback fined to the 3,781 students who received loop" in Fig. 1 from peer group affiliation to STR ratings and also successfully completed adolescent behavior) was not included in both portions of the self-report question- our analyses. naire, from which measures of parenting practices and adolescent behaviors were de- Method rived. This constituted 72% of the combined student bodies of the participating schools Sample and Procedures (excluding the California upperclassmen not Six public, 4-year high schools—three included in STR ratings). Of the remainder, in the Midwest and three on the West Coast 4% refused or were denied permission by —participated in the study. These included parents to participate in the study, 6% failed a small, rural school that drew its 400 stu- to fill out a questionnaire completely or dents (98% European-American) primarily credibly, and 18% were absent from class on from farm or working-class families; a school one or both questionnaire administration whose more ethnically diverse student body dates. The sample ranged in age from 14 to (only 70% of the 1,500 students were Euro- 19 (M = 15.5) and was fairly evenly divided pean-Americans) hailed from working- and by sex (52% female). Although a majority middle-class neighborhoods of a moderate- (61%) of the sample was European-Amer- sized city; a mid-sized school (1,000 stu- ican, there were also substantial numbers of dents) located in a pervasively European- African-Americans (12%), Asian-Americans American, upper-middle-class suburb; a (12%), and Hispanic youth (13%). large, ethnically diverse school (45% of the 2,500 students were African-American or Peer-rated Crowd Affiliation Hispanic) located in a predominantly upper- The Social Type Rating (STR) proce- middle-class community but also drawing dure (Brown, 1989) is an efficient mecha- from less economically privileged families nism for identifying adolescents' peer group in neighboring towns; a school whose 1,300 affiliation, based on their reputation among students were largely bifurcated between peers. It is somewhat similar to sociometric working- and upper-middle-class families, ratings by which younger children are clas- with Asian-American and Hispanic students sified into comparison groups (popular, re- collectively outnumbering European-Amer- jected, neglected, etc.; see, e.g., Terry & icans; and a large inner-city school with an Coie, 1991), except that the STR ratings take ethnically diverse student body—African- advantage of the more sophisticated social Americans formed the largest single racial/ system of adolescents in which there are ethnic group (40%)—from a broad range of commonly shared labels for particular peer socioeconomic backgrounds. Collectively, groups. the schools comprised a diverse sample in terms of size, location, socioeconomic back- Derived from earlier studies of adoles- ground, and ethnic composition. cent peer groups (Clasen & Brown, 1985; Poveda, 1975; Schweridinger & Schwen- All students in participating schools dinger, 1985), the STR procedure was a two- were invited to complete a self-report ques- step process. In the first step, school admin- tionnaire that focused on school-related istrators were asked to identify a set of boys behaviors (academic achievement, engage- and girls (within each ethnic group in multi- ment in classes, extracurricular participa- ethnic schools) in each grade who repre- Brown et al. 471 sented a good cross-section of the school's fact that few students are placed in the same sludent body. These students were inter- crowd by all raters; they also give the same viewed in small groups composed of stu- score to marginal crowd members (those dents of the same gender, same grade level, rated into a particular crowd by a bare major- and same ethnic group. T'hrough group dis- ity of raters) and central crowd members cussion, each group derived a list of the (those associated with the crowd by virtually school's major crowds; then each participant EJI raters). An alternative strategy, which was listed two boys and two girls in his or her employed for this study, is to create a series grade who were the leaders or most promi- of proportion scores representing the per- nent members of each crowd. From these centage of raters (excluding those who did lists a stratified sample was drawn in each not know a student well enough to assign to grade (stratified by crowd type, gender, and any crowd) who placed the student in a eliinicity, but with preference given to the given crowd. Thus, respondents received a most frequently listed students) to become score from 0 to 1.00 for each of their school's "STR raters" in the second step. Each rater, major crowds. Defining the outcome mea- accompanied by a friend of her or his own sures as continuous rather than categorical choosing, was individually interviewed. The variables also permitted us to use standard raters were presented with the list of crowds regression procedures for assessing the di- derived from the earlier group interviews, rect and mediated effects of our dependent thien asked to place each student in their measures (parenting behaviors). grade level into one of the crowds. Raters could indicate that they did not know a stu- Analyses in the present study were con- dent well enough to assign to a crowd. STR fined to proportion scores for six major ratings continued until each student had crowds: populars, jocks, brains, normals, been rated by at least 10 STR raters. Because druggies, and outcasts.^ These six crowds raters could only deal with about 300 names were widely recognized by raters in all par- in the time allotted for STR interviews, class ticipating schools. In fact, they comprised lists in the larger schools had to be parti- the bulk of STR crowd assignments (70%, on tioned and the number of raters increased to average). They are also the crowds on which ensure that all students received the re- most studies of adolescent peer groups have quired number of ratings. focused. Means, standard deviations, and in- tercorrelations among the crowd proportion There are several ways of using the STR scores are provided in Table 1. ratings to describe students' crowd affilia- tion. One is to create categorical assign- As would be expected with the scoring ments, that is, to place each student in the procedure we used, crowd proportion scores crowd to which a majority of STR raters as- were negatively skewed and negatively in- signed him or her. While useful for many tercorrelated.^ The comparatively low inter- analyses, categorical assignments mask the correlation between proportion scores for

^ The outcast group combined two crowds that were sometimes differentiated by STR raters: loners and nerds. Although distinctive in some ways, these groups share the same basic image among peers, especially regarding the behaviors of interest in this study: limited social skills, low self-image, low involvement in deviant behavior, and average or above average academic achievement levels. Thus, they appear to be different factions of the same major crowd (Brown, Lohr, & Trujillo, 1990). In fact, some STR raters confessed that calling someone a "loner" was basically a "nice" way of acknowledging that the person was a nerd. It may seem odd to consider outcasts (or at least loners) as a crowd because their label implies a lack of social interaction. Nevertheless, it is a widely recognized social category with a distinctive normative image (Brown et al., 1990) and pattern of peer pressures (Clasen & Brown, 1985). Thus, it clearly fits with our conceptual orientation toward crowds as reputation-based entities. ^ Crowd proportion scores are not independent; collectively, all of them (including those for crowd types not included in these analyses) must add to 1.00. Thus, as the number of crowd types increases the average score expected for each type diminishes. Crowd proportion scores in this study were low because of the substantial number of crowd types employed by STR raters. Further, as the proportion of assignments to a specific crowd increases, the proportion of assignments to other crowds must diminish; this explains the preponderance of negative correla- tions among crowd proportion scores. The negative skew of crowd proportion scores prompted us to repeat the major analyses using log transformations of the scores. Because results were substantively equivalent with both sets of scores we decided to retain the analyses based on the original scores. 472 Child Development TABLE 1

MEANS, STANDARD DEVIATIONS, AND INTEKCORKELATIONS OF CROWD PROPORTION SCORES

INTERCORRELATIONS

VAWABLE MEAN SD Popular Jock Brain Normal Druggie Popular .11 .20 Jock .07 .16 .02 Brain .09 .18 -.09 -.09 Normal .19 .23 -.11 -.10 -.07 Drusfiie .06 .16 -.11 -.08 -.14 -.14 Outcast . ... .18 .25 -.24 -.18 -.02 -.10 -.11

the populars and jocks and for the brains and styles. There is also some controversy about outcasts suggests that these pairs of crowds the best way to measure parenting behav- are more compatible and closely aligned iors: observations of parent-child interac- than other possible pairs; tbe relatively bigh tions (usually in contrived situations), par- negative correlation between proportion ents' self-report, or reports from children. scores for populars and outcasts suggests Each approach has strengths and limitations; that this pair of crowds is quite incompati- observations lack ecological validity, and ble. Previous research on adolescent crowds botb parents and children seem to distort re- confirms these patterns (see Brown, Mory, & ality somewhat in their reports of parenting Kinney, in press), thus bolstering the valid- behavior. Because of the size of our sample, ity of the STR procedure. we were forced to rely exclusively on ado- lescents' report of parenting behaviors. To Other Measures minimize distortions in their responses, we Apart from crowd affiliation (which was focused on measures of concrete parenting based on peer ratings), all other measures behaviors, ratber than asking respondents to were derived from students' responses to make more general (subjective) judgments self-report questionnaire items. These mea- about their parents' child-rearing practices. sures fell into three basic sets: demograpbic characteristics, respondents' report of par- From the wider set of variables on the ents' child-rearing practices, and scores on questionnaire relating to parent attitudes three measures of adolescent behavior. Each and behaviors, we selected several that met of these sets is described below. our operational criteria (i.e., assessed con- crete behaviors) and defined constructs that Demographic characteristics.—Stu- were especially pertinent to the questions dents provided information on their back- on which this study focused. The first con- ground and current family situation. Spe- struct was parental emphasis on achieve- cifically, they indicated their sex, ethnic ment. Included in this 15-item measure affiliation, family structure (classified as in- were answers to the following questions: tact, single-parent, stepparent, or other), and "How often does each of your parents (a) family socioeconomic status (SES), based on help you with homework when asked, (b) parents' level of education. know how you're doing in school, (c) go to Parenting practices.—Investigators school programs for parents, (d) watch you have taken a variety of approaches to mea- in sports or activities, and (e) help you in suring parenting bebaviors. Some have at- choosing your courses?"; "What is the low- tempted to identify global parenting styles est semester grade you could get in each of that reflect an adult's child-rearing values these subject areas witbout your parents get- and orientations (Baumrind, 1991a). Bron- ting upset? (a) English, {b) math, (c) science, fenbrenner (1991), however, has cautioned (d) social studies"; "How important is it to that a given parenting style is not necessarily your parents or guardians that you work hard translated into the same behaviors by all par- on your schoolwork?" All items were an- ents. Thus, in view of our model, we chose swered on a 5-point Likert scale. The items to measure parenting practices directly formed a scale with acceptable internal con- rather than inferring them from parenting sistency (alpha = .84); scores represented Brown et al. 473 the mean of item responses (1 = low, 5 = (Donovan & Jessor, 1985; Dornbusch et al., 3 1987). Respondents also completed a five- item drug use scale (alpha = .86), indicating, Students also completed a five-item on a four-point scale (never, once or twice, pjirental monitoring scale simileir to ones several times, often), how often in the past that have been used in previous studies 6 months they had used various controlled (Dishion, 1990; Dombusch et al., 1985; Pat- substances. Scale scores (the mean of item terson & Stouthamer-Loeber, 1984). On a responses) ranged from 1.00 (low) to 4.00 three-point scale (don't know, know a little, (high). Similar measures in previous studies know a lot) they indicated how much their have been sbown to be valid indicators of psirents really knew about who their friends adolescent drug use (Needle, McCubbin, were, how they spent their money, where Lorence, & Hochhauser, 1983; O'Malley, they were after school, where they went at Bachman, & Johnston, 1983; Polich, 1982). night, and what they did with their free time. Finally, respondents completed the 10-item The scale had an internal consistency alpha self-reliance subscale from Greenberger's of .80; scale scores (mean item score) ranged Psychosocial Maturity Inventory (Form D; from 1.00 (low) to 3.00 (high). Greenberger, Josselson, Knerr, & Knerr, The third measure indicated the degree 1974). This scale measures adolescents' feel- to which parents engaged their child in Joint ings of internal control and the ability to decision making rather than making unilat- make decisions witbout extreme reliance eral decisions for the child or allowing the upon others. Items are answered on a four- child to decide things completely by himself point Likert scale. Like the drug use mea- or herself. This measure was based on a 13- sure, scale scores (the mean of item re- item scale in which respondents indicated, sponses) ranged from 1.00 (low) to 4.00 on a 5-point scale (from "my parents decide (high). In this sample the scale's alpha was this without discussing it with me" to "I de- .81. cide this without discussing it with my par- ents"), how they and their parents arrived at a decision about a variety of issues com- Plan of Analyses monly faced by high school students: school Our conceptual model postulated that classes, curfew times, spending patterns, use adolescent behaviors would mediate the as- of alcohol, and so on (Dornbusch et al., 1985; sociation between parenting practices and Steinberg, 1987). The scale's alpha was .82. crowd affiliation. Analyses followed a stan- Scale scores, ranging from 0 to 1.00, indi- dard procedure for testing sucb mediating cated the proportion of items on which deci- effects (Baron & Kenny, 1986). First, we ex- sions were derived jointly as opposed to uni- amined the direct association between par- laterally (by parents only or the adolescent enting practices and crowd affiliation. Then, only).* we determined whether the significant di- rect effects of parenting practices dimin- Adolescent behaviors.—Three mea- ished or disappeared when adolescent be- sures assessed adolescent behaviors that baviors were introduced as mediating were expected to be related to parenting variables. Following this, we examined asso- practices. Self-reported grade-point average ciations between parenting practices and ad- (GPA), scored on the standard 4-point scale, olescent behaviors and between adolescent was employed as a measure of academic behaviors and crowd affiliation. Separate achievement. Self-reported grades have analyses were conducted for each crowd af- been found to correlate highly (r = .75) with filiation score, using multiple regression actual grades taken from student transcripts techniques.

^ Several other items were considered for this scale but dropped on the basis of initial factor analyses and reliability analyses. The omitted items seemed to focus on parental strategies in response to poor academic performance (e.g., "How often does your parent make sure you do your homework") rather than behaviors that underscore the value they attach to achievement. * Previous studies have emphasized j'oini versus unilateral decision-making, without differ- entiating whether unilateral decisions are made by parents or the child (Yee & Flanagan, 1985). Those who have differentiated the two types of unilateral decisions have reported curvilinear effects, in which adolescent outcomes were highest in families practicing joint decision-making and not significantly different in femilies practicing "youth alone" or "parent alone" strategies (Dornbusch et al., 1985, 1989). Thus, the most appropriate usage of this measure for our analyses seems to be to calculate the proportion of joint decisions. 474 Child Development Results these demographic differences corre- Characteristics of Each Set of Variables sponded to ethnographers' observations Table 2 presents the mean scores, stan- about the crowds; ethnographers have por- dard deviations, and intercorrelations for the trayed the druggies as dominated by lower- measures of parenting practices and adoles- class youth. Collectively, demographic vari- cent behaviors. Within each set of measures ables accounted for between 7.3% and 13% correlations were positive but relatively low, of the variance in crowd affiliation scores suggesting (as expected) that the measures (see Table 4). Because of tbese differences, addressed distinctive aspects of parenting or we included demographic variables as con- adolescent behavior. trol variables in all subsequent analyses. To determine the strength of association Direct Associations between Parenting between student background cbaracteristics Practices and Crowd Affiliation and crowd affiliation, we correlated crowd The first step in testing the conceptual affiliation scores with SES and also con- model was to examine the direct effects of ducted a series of ANGOVAs (covarying for parenting practices on crowd affiliation. This SES) that examined the independent effects was accomplished by regressing each crowd of sex, grade level, ethnicity, family struc- affiliation score on each parenting variable, ture, and school on each crowd affiliation after controlling for effects of the demo- score. Results are summarized in Table 3. graphic variables noted above as well as the SES was positively correlated with member- other two parenting variables. Tbe paths in ship in the popular (r = .13) and brain (r = these analyses that were significant at p < .14) crowds but virtually uncorrelated (r's < .05 or better are reported in Figure 2. Paren- .05) witb other crowd scores. There were tal emphasis on achievement was positively significant sex differences in scores for all associated with membership in the popular, crowds; girls had stronger associations than jock, and normal crowds and negatively as- boys with the popular, brain, and normal sociated with membership in the druggies. crowds, but weaker associations with the Parent monitoring was positively associated jocks, druggies, and outcasts. Grade differ- with membership in the brain crowd and ences were significant for half of the crowds negatively associated with membership in (jocks, brains, and outcasts), although there the druggies. Joint decision making was pos- was no clear or consistent pattern to these itively associated with membership in the differences. Being from an intact family brains and normals and negatively associ- seemed to enhance membersbip in the brain ated with membersbip in tbe druggies. crowd and diminish membersbip in tbe Tbere were no significant patbs between druggie crowd. All crowd affiliation scores parenting variables and membersbip in the differed significantly by scbool, suggesting outcast crowd. that there were substantial school- or com- Adolescent Behaviors as Mediating munity-based differences in ttle size of eacb Variables crowd. Except for the absence of association The next step in testing the conceptual between SES and druggie crowd scores. model was to determine if the effects of par-

TABLE 2 MEANS, STANDARD DEVIATIONS, AND INTERCORRELATIONS OF PARENTING AND PSYCHOSOCIAL ADJUSTMENT VARIABLES

INTERCORRELATIONS

VARIABLE MEAN SD PEA Mon JDM CPA Drug Use Parental emphasis on achievement (PEA) .. 3.12 .80 Monitoring (Mon) 2.31 .49 .33* Joint decision making (JDM) 49 .25 .34* .36* CPA 2.76 .85 .26* .17* .17* Drug use 1.54 .72 -.12* -.27* -.20* -.27* Self-reliance 3.04 .53 .13* .10 .02 .19* -.07* *p< .001. Brown et al. 475 TABLE 3

DEMOGRAPHIC DIFFERENCES IN MEAN CROWD AFFILIATION SCORES

CROWD AFFILIATION SCORE DEMOGRAPHIC GROUP Popular Jock Brain Normal Druggie Outcast Sex: Male 08 .09 .08 .16 .07 .22 Female 16 .03 .10 .21 .05 .15 Crade level: Freshmen 12 .05 .07 .19 .06 .22 Sophomores 12 .07 .10 .19 .06 .15 Juniors 11 .06 .09 .21 .07 .17 Seniors 14 .06 .10 .19 .04 .18 Family structure: Intact 12 .07 .10 .20 .05 .19 Single parent 11 .06 .08 .18 .07 .18 Step-parent 14 .04 .06 .19 .09 .16 Other 10 .04 .06 .16 .10 .16 Ethnicity: African-American 09 .08 .05 .19 .00 .11 Euro-American 14 .06 .09 .20 .09 .14 Asian-American 10 .03 .18 .13 .02 .21 Hispanic 07 .04 .05 .18 .02 .12 School: A 11 .10 .13 .24 .06 .13 B 22 .10 .09 .12 .05 .22 C 11 .06 .06 .25 .06 .14 D 06 .06 .05 .19 .06 .19 E 09 .02 .10 .16 .06 .13 F 14 .05 .10 .18 .06 .23 NOTE.—Scores for each variable are adjusted for the effects of all other variables plus socioeconomic status. Schools are listed in the order in which they are described in the Method section. Students who did not belong to any of the four ethnic comparison groups were omitted from these analyses. Further details of the analyses are available from the first author. enting practices on crowd affiliation were significant at an equivalent level of magni- mediated by adolescent behaviors. Of the tude when adolescent bebaviors were in- nine initially significant paths between par- cluded in the regression analyses (see Fig. enting practices and crowd affiliation (Fig. 3). Thus, as expected, the effects of parent- 2), five were reduced to nonsignificance, ing practices were largely mediated by ado- three were reduced substantially in magni- lescent behaviors. tude, and only one (the association between parental emphasis on achievement and Collectively, parenting practices and membership in the jock crowd) remained adolescent behaviors accounted for between

TABLE 4 PROPORTION OF VARIANCE IN CROWD AFFILIATION SCORES ACCOUNTED FOR BY BACKGROUND VARIABLES, PARENTING PRACTICES, AND ADOLESCENT BEHAVIORS

Background Parenting PP + Adolescent Crowd Variables Practices (PP) Behaviors Tota Populars .129 .009 .033 .162 Jocks .073 .013 .016 .089 Brains .088 .010 .139 .227 Normals .070 .004 .007 .077 Druggies .053 .026 .175 .230 Outcasts .087 .001 .023 .111 476 Child Development

Parental emphasis on achievement

Parent monitoring

Joint decision- making

FIG. 2.—Observed direct associations between parenting practices and crowd affiliation. Standard- ized betas are shown for all paths significant at at least the .05 level. *p < .05; **p < .01; ***p < .001.

2% and 15% of the variance in crowd affilia- practices and adolescent behaviors and be- tion scores. Except for the jock and normal tween adolescent behaviors and crowd affil- scores, this was a notable improvement over iation. Results of these analyses are depicted the initial regression analyses assessing the in Figure 3, which shows all paths that were direct effects of parenting practices on significant at p < .05 or better. crowd affiliation (see Table 4). First, to assess the direct, independent To further establish the mediating role effect of each parenting practice on each ad- of adolescent behaviors in our model, we ex- olescent bebavior, tbe score for each of the amined the associations between parenting three adolescent behavior variables was re-

.122"' Popuiar

.108'" Parentai emphasis Grade average on achievement Jock

Brain Parent monitoring i3njguse

Nomial

Joint decision- Self-reiiance making Druggie •.0B3'

Outcast

FIG. 3.—Observed associations between parenting practices, adolescent behaviors, and crowd affiliation. Standardized betas are shown for all paths significant at at least the .05 level. Dotted lines indicate curvilinear effects. *p < .05; **p < .01; ***p < .001. Brown et al, 477 gressed, in turn, on scores on the other two which served to confirm ethnographers' ob- variables and the parenting measures (as servations that each crowd features a distinc- well as demographic control variables). Pa- tive life-style. rental emphasis on achievement was posi- tively related to adolescents' reports of their Summory self-reliance and grade-point average. Pa- Collectively, these analyses provided rental monitoring was inversely associated clear evidence of the mediating role that ad- with drug use and directly associated with olescent behaviors play between parenting self-^reliance. Also, higher rates of joint deci- practices and adolescent crowd affiliation. sion making were predictive of higher stu- Parenting practices had a significant impact dent grade-point averages and lower levels on adolescent behaviors, which in turn were of drug use and self-reliance. These findings significantly associated with crowd affilia- w£;re consistent with results of previous tion, sucb tbat each crowd score had a studies except for (o) the tendency of joint unique profile on these three behavioral decision making to depress, rather than en- measures. Even with the inclusion of adoles- hance, self-reliance, and {b) the absence of cent behaviors as mediating variables, how- a significant association between parent ever, some direct associations between par- monitoring and student acbievement. enting behaviors and crowd affiliation scores remained significant. Finally, regression analyses examined Crowd affiliation scores also were mod- the relation between adolescent behaviors estly associated with adolescents' back- and peer crowd affiliation. In most cases, ground characteristics, some of which (SES, previous studies have suggested that crowd family structure) reflected features of their affiliation is directly enhanced or dimin- family background. The ability of these ished by the behaviors we considered. In a three sets of variables—parenting practices, few instances, however, the relation has adolescent behaviors, and background char- been depicted as curvilinear, that is, jocks acteristics—to account for variance in crowd and normals have been depicted as moder- affiliation scores differed considerably ate academic achievers, and brains and nor- among the crowds. In most cases, the direct mals have been portrayed as harboring mod- and indirect effects of parenting practices erate levels of self-esteem or self-reliance. superseded the effects of background char- For these four instances, we tested for possi- acteristics, but in all cases these variables ble curvilinear as well as linear effects in the explained a relatively modest proportion of regression analyses. With few exceptions, variance in crowd affiliation scores. the significant paths for these analyses, shown in Figure 3, confirmed the behavioral Discussion profiles of crowds that had been observed in previous studies. Membership in the popu- Researchers have clearly documented lar crowd was predicted by high scores on the powerful role parents play prior to ado- all three behaviors, whereas membership in lescence in shaping children's social skills the outcasts was predicted by relatively low and directing their relationships with peers scores on these behaviors (the tendency for (Ladd, Muth, & Hart, in press; Parke & outcasts to have low GPAs was unexpected). Bhavnagri, 1989; Putallaz, 1987); studies of Meirlbership in the brains was related to how parental infiuences in this area con- high grades and low drug use but not to self- tinue through adolescence are much less ex- reliance (a curvilinear effect had been pre- tensive. By linking two sets of findings that dicted). Membership in the druggies was as- have been pursued ratber independently— sociated with low grades and high drug use; examining the infiuence of parenting prac- the expected positive association with self- tices on adolescent behavior patterns, and reliance was not significant. Academic showing how these bebaviors differentiate achievement was the only behavior that pre- members of various adolescent crowds—the dicted membership in tbe jock crowd; as ex- study joins a number of investigations dem- pected, the relation was curvilinear. And as onstrating that although parental influence expected, drug use discouraged member- over peer associations may diminish in ado- ship in the normals, whereas moderate GPA lescence, it is far from inconsequential. levels enhanced it, but the expected curvi- Investigators have acknowledged that linear association with self-reliance was not parents retain some control over their ado- significant for this crowd. Each crowd affili- lescent child's choice of peer associates ation score had a unique pattern of associa- through their selection of the neighborhood tion With the adolescent bebavior measures. in which the family lives or the schools. 478 Child Development churches, and community organizations in tions of ethnic groups: European-Americans which family members participate (Rubin & were overrepresented among the high- Sloman, 1984). Our study confirmed this in status, popular crowd; Asian-Americans were the substantial differences among schools in overrepresented among brains; and African- the size of various crowds. In this respect, American students were overrepresented in our findings relate to the "community stud- the jock crowd. Whether or not these pat- ies" of several decades ago, in which the au- terns reflected ethnic differences in parental thors contended that the high school peer expectations or family reward structures is group system reflected characteristics of'the an important area for further research. community in which it was located (Cole- man, 1961; Hollingshead, 1949; Lynd & Family structural characteristics such as Lynd, 1957). Often, however, these studies socioeconomic status, ethnic background, asserted that peer group membership was and marital arrangements reflect one im- primarily a function of socioeconomic divi- portant way in which parents retain some sions within the community, whereas in our influence over teenagers' peer group associ- study socioeconomic status was a relatively ations, but our model was concerned primar- poor predictor of crowd affiliation—and ily with process characteristics, namely, par- community (school) differences in crowd af- enting practices. Our confidence in the filiation scores remained highly significant causal relationships specified in the model after controlling for students' socioeconomic are tempered by the fact that the findings background. These school-based differences are not based on longitudinal data. Hays and remind us of the limitations of generalizing Revetto (1990) demonstrated that several characteristics of the adolescent peer group causal arrangements can be supported in system across contexts. They also under- correlational data sucb as ours. There is, score the need to consider contextual fea- however, a growing body of research evi- tures when examining such questions as dence, including some longitudinal work, how parents influence teenagers' peer group that supports the causal arrangement of vari- affiliations. Parents' efforts to direct a child ables specified in our model. Feldman and toward an athletically oriented crowd, for ex- Wentzel (1990) found that harsh discipline ample, may be fruitless if such a crowd is strategies affected early adolescent boys' so- virtually nonexistent—as appeared to be the ciometric status indirectly, tbrough the boys' case in school E (see Table 3). capacity to inbibit aggressive behavior. Whitbeck, Simons, Conger, and Lorenz Altbough ethnographers have empha- (1989) found tbat parental values influenced sized family socioeconomic status as an im- the values of their teenage offspring, which, portant factor in adolescents' peer group af- in turn, were significantly related to the filiations (Eckert, 1989; Hollingshead, 1949; teenagers' associations with deviant peers. Larkin, 1979), we found other family back- Working with younger age groups (fourth ground characteristics—family structure and through seventh graders), Dishion (1990; ethnicity—to be more salient. Students from Dishion, Patterson, Stoolmiller, & Skinner, intact families appeared to be overrepre- 1991) has reported both correlational and sented among the brains and underrepre- longitudinal evidence that parenting strate- sented among druggies, which was precisely gies (discipline and monitoring) predict the opposite pattern of students from step- child behaviors (academic performance and parent families, who were also underrepre- deviant activity), which in turn affect pat- sented in the jock crowd. It may be that the terns of peer associations (sociometric status hostility that many teenagers harbor toward and involvement with deviant peers). Our a stepparent (Fine, 1986) "spills over" to causal ordering is also consistent with other their relationship with school adults, thus longitudinal evidence that deviant behavior constraining their ability to thrive in peer predicts association with deviant peers groups (such as jocks and brains) tbat feature rather than the reverse (Magnusson, 1988). close relationships with school staff. This model contrasts sharply with ones The predominance of European-Amer- proposing that parenting strategies affect icans in the druggie crowd was consistent peer group affiliations directly (Elliott et al., with research evidence indicating that Euro- 1985; Kandel & Andrews, 1987). The conse- pean-American teenagers are more involved quences of these different perspectives are in drug use than adolescents from other well illustrated by our findings witb refer- ethnic groups (McCord, 1990). In other ence to the popular crowd. Had we applied respects, ethnic differences in crowd affil- a "direct effects" perspective to our data we iation scores paralleled stereotypic orienta- might have argued that parents may Brown et al, 479 "nudge" their child toward the popular and association with prosocial peer groups. crowd by encouraging academic achieve- Yet, in view of the amount of variance in ment, a bebavior furtber enbanced by popu- crowd affiliation scores accounted for by our lar crowd norms (considering the relatively parenting variables, it may appear as if par- high achievement levels of this crowd). Yet, ents exert some modest influence over their in view of the absence of direct paths be- child's associations with brains or druggies, tween otber parenting behaviors and popu- but that otherwise parental influences on lar crowd affiliation, we probably would crowd affiliation are trivial. It may well be have concluded that peer group norms, that parents have an impact on adolescents' rather than parenting bebaviors, must be re- affiliations with certain crowds more so than sponsible for populars' propensity to engage others, but, for several reasons, conclusions in drug use. By contrast, our "mediated about the magnitude of parental influences model" suggests that parents have a hand seem premature. First, a skewed distribu- in shaping both the prosocial and antisocial tion of scores on outcome measures (as was behavior patterns that characterize the popu- tbe case here) typically depresses the lar crowd. Parental emphasis on achieve- amount of variance accounted for, so that ment inspires higher academic performance, parenting effects may be underestimated in which fosters association with the populars. our results. Second, adolescent crowds are Yet, when the emphasis on achievement is differentiated on a much wider range of be- accompanied by efforts to monitor the ; haviors, which in turn are likely to be af- child's behavior, the child is oriented more fected by a mucb broader array of parenting tovi'ard the brain crowd, whereas when par- practices, than were examined in this study. ents fail to monitor the child and do not en- Indeed, the variables we examined seem courage joint decision making, the child is highly salient for the brains (academic per- more likely to become associated with the formance, parental empbasis on achieve- popular crowd, whose desirable behaviors ment) and druggies (drug use, parental mon- (academic achievement) are accompanied itoring), which probably accounts for the fact by undesirable ones (drug use). that we explained more of the variance in scores for these crowds than other crowds. In fact, our model proposes tbat peer groups are unlikely to countermand parental Most important, our study was confined influences—for example, to lead an adoles- to measures of parenting practices and chil- cent into deviant activities despite parents' dren's behavior in adolescence, whereas it is efforts to orient the child toward prosocial likely that parental influences on adolescent values and behaviors. Instead, peer group crowd affiliation operate on a more extended norms serve primarily to reinforce behaviors developmental timetable. Cairns et al. and predispositions to which parents (1988) bave sbown tbat parenting character- (through parenting strategies and/or family istics contribute to the division of youngsters background characteristics) have already by middle childhood into clusters of deviant contributed. Although this contrasts with and nondeviant peers, which may be the some traditional perspectives on deviance forerunners of prosocial (e.g., brains, jocks) (Sutherland & Cressey, 1978), it is consistent and antisocial (e.g., druggies) crowds in ado- with some more recent work in this area lescence. Thus, a teenager's crowd affilia- (e.g., Dishion et al., 1991; Snyder, Dishion, tion may be influenced by parenting behav- & Patterson, 1986). Of course, support for iors many years prior to adolescence. Snyder these suppositions awaits longitudinal data et al. (1986) found evidence that associations that examine the full conceptual model dis- with deviant peers in adolescence seemed played in Figure 1, including the "feed- to be tbe result of a series of events across back" loop of crowd norms and peer pres- childhood and early adolescence. What is sures. This study was simply a modest first more, the parenting behaviors that contrib- step in this process, but the findings should uted to this chain of events changed with encourage researchers to pursue the full age: in childhood, parents' effectiveness in model in earnest. teaching social skills and disciplining antiso- The findings make explicit what re- cial bebavior were key factors; by early ado- mains implicit in much of the research on lescence, parental monitoring of tbe child's deviant bebavior: tbat whereas inadequate behavior was more salient. In other words, or undesirable parenting practices may con- our data did not consider the cumulative or tribute to the child's involvement in deviant lagged effects of parenting behaviors prior peer groups and deviant activities, positive to adolescence, wbicb may bave as much if parenting fosters desirable behavior patterns not more impact (as parenting practices in 480 Child Development adolescence) on nurturing behaviors that commentary. In S. Lamborn (Chair), Authori- help direct teenagers to particular peer tative parenting and adolescent develop- groups. Researchers must now turn attention ment. Symposium presented at the biennial to the task of tracing the linkages between meetings of the Society for Research in Child parents' efforts to monitor peer associates Development, Seattle. and foster social skills in childhood, the for- Brown, B. B. (1989). Social Type Rating Manual. mation and metamorphosis of preadolescent Madison: National Center on Effective Sec- cliques, and the emergence of adolescent ondary Schools, University of Wisconsin— peer groups and processes whereby teen- Madison. agers become associated with a particular Brown, B. B. (1990). Peer groups and peer cul- crowd. tures. In S. S. Feldman & C. R. Elliott (Eds.), At the threshold: The developing adolescent There may be other facets of teenage (pp. 171-196). Cambridge, MA: Harvard Uni- peer relations in which parental influences versity Press. operate rather directly, but as for adoles- Brown, B. B., Lohr, M. J., & Trujillo, C. (1990). cents' crowd affiliations, tbe results of tbis Multiple crowds and multiple life-styles: Ad- study suggest tbat parental influences are olescents' perceptions of peer group stereo- largely indirect, through their impact on be- types. In R. E. Muuss (Ed.), Adolescent be- haviors by which teenagers become associ- havior and society, 4th ed. (pp. 30-36). New ated witb a particular crowd. What is more, York: McGraw-Hill. parents have the capacity to direct their child toward an essentially prosocial crowd Brown, B. B., Mory, M., & Kinney, D. (in press). (such as the brains) or an antisocial crowd Casting adolescent crowds in a relational per- (sucb as tbe druggies) or a peer group with spective: Caricature, channel, and context. In mixed characteristics (such as populars or R. Montemayor, C. R. Adams, & T. P. Cullotta outcasts). To be sure, parenting practices (Eds.), Advances in adolescent development: and family background cbaracteristics can- Vol. 6. Personal relationships during adoles- not determine a teenager's crowd affiliation, cence. Newbury Park, CA: Sage. but their influence should not be dis- Buff, S. A. (1970). Creasers, dupers, and hippies: counted. In view of this, researchers are en- Three responses to the adult world. In L. couraged to pay more attention to the Joint Howe (Ed.), The white majority (pp. 60-77). influences of parents and peers on adoles- New York: Random House. cent bebavior. Cairns, R. B., Cairns, B. D., Neckerman, H. J., Gest, S. D., & Gariepy, J.-L. (1988). Social networks and aggressive behavior: Peer sup- Referenees port or peer rejection? Developmental Psy- Baron, R. M., & Kenny, D. A. (1986). The modera- chology, 24, 815-823. tor-mediator variable distinction in social psy- Clasen, D. R., & Brown, B. B. (1985). The multidi- chological research: Conceptual, strategic, mensionality of peer pressure in adolescence. and statistical considerations. Journal of Per- Journal of Youth and Adolescence, 14, 451- sonality and Social Psychology, 51, 1173- 468. 1182. Coleman, J. S. (1961). The adolescent society. Baumrind, D. (1991a). Parenting styles and ado- New York: Free Press. lescent development. In H. M. Lerner, A. C. Coombs, R. H., & Landsverk, J. (1988). Parenting Petersen, & J. Brooks-Cunn (Eds.), Encyclo- styles and substance use during childhood pedia of adolescence (Vol. 2, pp. 746-758). and adolescence. JournoZ of Marriage and the New York: Carland. Family, 50, 473-482. Baumrind, D. (1991b). The influence of parenting Dishion, T. J. (1990). The peer context of trouble- style on adolescent competence and sub- some child and adolescent behavior. In P. E. stance use. Journal of Early Adolescence, 11, Leone (Ed.), Understanding troubled and 56-95. troubling youth (pp. 128-153). Newbury Berndt. T. J. (1979). Developmental changes in Park, CA: Sage. conformity to peers and parents. Develop- Dishion, T. J., Patterson, G. R., Stoolmiller, M., mental Psychology, 15, 606—616. & Skinner, M. L. (1991). Family, school, and Bowerman, C. E., & Kinch, J. W. (1959). Changes behavioral antecedents of early adolescent in family and peer orientation between the involvement with antisocial peers. Develop- 4th and 10th grades. Social Forces, 37, 206- mental Psychology, 27, 172-180. 211. Donovan, J., & Jessor, R. (1985). Structure of prob- Brittain, C. V. (1963). Adolescent choices and par- lem behavior in adolescence and young adult- ent-peer cross-pressures. American Sociologi- hood. Journal of Consulting and Clinical Psy- cal Review, 28, 385-391. chology, 53, 890-904. Bronfenbrenner, U. (1991, April). Discussant's Dombusch, S. M., Carlsmith, J. M., Bushwall, P. Brown et al. 481

L., Ritter, P. L., Leiderman, P. H., Hastorf, Hays, R. D., & Revetto, J. P. (1990). Peer cluster A. H., & Gross, R. T. (1985). Single parents, theory and adolescent drug use: A reanalysis. extended households, and the control of ado- Journal of Drug Education, 20, 191-198. lescents. Child Development, 56, 326-341. Hollingshead, A. B. (1949). Elmtown's youth: The Dornbusch, S. M., & Ritter, P. L. (1991, April). impact of social class on adolescents. New Fam,ily decision-making and authoritative York: Wiley. parenting. Paper presented at the biennial Kandel, D. B., & Andrews, K. (1987). Processes of meetings of the Society for Research in Child adolescent socialization by parents and peers. Development, Seattle. International Journal of the Addictions, 22, Dornbusch, S. M., Ritter, P. L., Leiderman, P. H., 319-342. Roberts, D. F., & Fraleigh, M. J. (1987). The Ladd, G. W., Muth, S., & Hart, G. H. (in press). relation of parenting style to adolescent Parents' management of children's peer rela- school performance. Child Development, 58, tions: Facilitating and supervising children's 1244-1257. activities in the peer culture. In R. D. Parke Eckert, P. (1989). Jocks and burnouts: Social cate- & G. W. Ladd (Eds.), Family-peer relations: gories and identity in the high school. New Modes of linkage. Hillsdale, NJ: Erlbaum. York: Teachers College Press. Larkin, R. W. (1979). Suburban youth in cultural Eder, D. (1985). The cycle of popularity: Interper- crisis. New York: Oxford. sonal relations among female adolescence. Loeber, R., & Dishion, T. J. (1983). Early pre- Sociology of Education, 58, 154-165. dictors of male adolescent delinquency: A re- Elliott, D. S., Huisinga, D., & Ageton, S. S. (1985). view. Psychological Bulletin, 94, 68-99. Explaining delinquency and drug use. Bev- Lynd, R. S., & Lynd, H. M. (1957). Middletown: erly Hills, CA: Sage. A study in American culture. New York: Har- Featherman, D. L. (1980). Schooling and occupa- court. Brace, and World. tional careers: Constancy and change in Magnusson, D. (1988). Individual development worldly success. In O. Brim & J. Kagan (Eds.), from, an interactional perspective. Hillsdale, Constancy and change in human develop- NJ: Erlbaum. ment (pp. 675-738). Cambridge, MA: Har- Marcos, A. G., Bahr, S. J., & Johnson, R. E. (1986). vard University Press. Test of a bonding/association theory of ado- Feldman, S. S., & Weritzel, K. R. (1990). The rela- lescent drug use. Social Forces, 65, 135-161. tionship between parenting styles, son's self- Massey, J. L., & Krohn, M. D. (1986). A longitudi- restraint, and peer relations in early adoles- nal examination of an integrated social pro- cence. Journal of Early Adolescence, 10, cess model of deviant behavior. Social 439-454. Forces, 65, 106-134. Fine, M. (1986). Perceptions of stepparents: Varia- McCord, J. (1990). Problem behaviors. In S. S. tion in stereotypes as a function of current Feldman & G. R. Elliott (Eds.), At the thresh- family structure. JournaZ of Marriage and the old: The developing adolescent (pp. 414— Family, 48, 537-543. 430). Cambridge, MA: Harvard University Floyd, H. H., & South, D. R. (1972). Dilemma of Press. youth: The choice of parents or peers as a Needle, R., McCubbin, H., Lorence, J., & Hoch- frame of reference for behavior. Journal of hauser, M. (1983). Reliability and validity of Marriage and the Family, 34, 627-634. adolescent self-reported drug use in a family- Cecas, V., & Seff, M. A. (1990). Families and ado- based study: A methodological report. Inter- lescents: A review of the 1980s. Journal of national Journal of the Addictions, 18, Marriage and the Family, 52, 941-958. 901-912. Clynn, T. J. (1984). Adolescent drug use and the Oetting, E. R., & Beauvais, F. (1987). Peer cluster family environment: A review. Journal of theory, socialization characteristics, and ado- Drug Issues, 14, 271-295. lescent drug use: A path analysis. Journal of Gieenberger, E., Josselson, R., Knerr, C, & Knerr, Counseling Psychology, 34, 205-213. B. (1974). The measurement and structure of O'Malley, P. M., Bachman, J. G., & Johnston, L. psychosocial maturity. Journal of Youth and D. (1983). Reliability and consistency in self- Adolescence, 4, 127-143. reports of drug use. International Journal of Grotevant, H., & Cooper, C. R. (1986). Individua- the Addictions, 18, 805-824. tion in family relationships: A perspective on Parke, R., & Bhavnagri, N. P. (1989). Parents as individual differences in the development of managers of children's peer relationships. In identity and role-taking skill in adolescence. D. Belle (Ed.), Children's social networks Human Development, 29, 82-100. and social support (pp. 241—259). New York: Hauser, S. T., Powers, S., Noam, G., Jacobsen, A., Wiley. Weiss, B., & Follansbee, D. (1984). Familial Parsons, T. (1942). Age and sex in the social struc- contexts of adolescent ego development. ture of the United States. American Sociologi- Child Development, 55, 195-213. cal Review, 7, 604-616. 482 Child Development Patterson, G. R., & Stouthamer-Loeber, M. (1984). Snyder, J., Dishion, T. J., & Patterson, G. R. The correlation of family management prac- (1986). Determinants and consequences of as- tices and delinquency. Child Development, sociating with deviant peers during preado- 55, 1299-1307. lescence and adolescence. Journal of Early Polich, J. M. (1982). The validity of self-reports in Adolescence, 6, 20-34. alcoholism research. Addictive Behaviors, 7, Steinberg, L. (1987). Single parents, step-parents, 123-132. and the susceptibility of adolescents to anti- Poveda, T. G. (1975). Reputation and the adoles- social peer pressure. Child Development, 57, cent girl: An analysis. Adolescence, 37, 127- 269-275. 136. Sutherland, E. W., & Cressey, D. R. (1978). Crimi- Putallaz, M. (1987). Maternal behavior and chil- nology (10th ed.). Philadelphia: Lippincott. dren's sociometric status. Child Develop- Terry, R., & Coie, J. D. (1991). A comparison of ment, 58, 324-340. methods for defining sociometric status Riesman, D. (1961). The lonely crowd. New Ha- among cbildren. Developmental Psychology, ven, CT: Yale University Press. 27, 867-880. Rubin, Z., & Sloman, J. (1984). How parents in- Whitbeck, L. B., Simons, R. L., Conger, R. D., & fluence tbeir children's friendships. In M. Lorenz, F. O. (1989). Value socialization and Lewis (Ed.), Beyond the dyad (pp. 115-140). peer group affiliation among early adoles- New York: Plenum. cents. Journal of Early Adolescence, 9, 436— Schwendinger, H., & Schwendinger, J. S. (1985). 453. Adolescent subcultures and delinquency. Yee, D. K., & Flanagan, C. (1985). Family environ- New York: Praeger. ments and self-consciousness in early adoles- Seginer, R. (1983). Parents' educational expecta- cence. Journal of Early Adolescence, 5, 59- tions and children's academic achievements: 68. A literature review. Merrill-Palmer Quar- terly, 29, 1-23.