Aggression and delinquency: Family and environmental factors

Journal of Youth and Adolescence, VoL 17, No. 3, 1988

Aggression and Delinquency: Family and Environmental Factors Lisabeth Fisher DiLalla, ~.s Christina M. Mitchell, z Michael W. Arthur, 3 and Pauline M. Pagliocca'

Received September 29, 1987; acceptedDecember 2, 1987

Juvenile delinquency has become an increasing concern to society; aggressive behaviors are particularly harmful. This study examined parent and youth behaviors and personality types that may influence delinquent and aggressive behaviors. Youths were referred by the court to an intervention program; ratings of delinquency and aggression were derivedfrom parent reports, self-reports, and court referral data. Results showed that high parent ratings o f youth aggressiveness were related to high turmoil in the home and to youths" positive opinions o f delinquent peers, while high aggressiveness o f the youths" referring offenses was related to lax punishment. Developmentally, this suggests that in adolescence both the peer group and home influences are important in shaping different aspects o f the youths" aggressive and delinquent behaviors. This study was partially conducted under University of Virginia Research Policy Council Grant No. 199505. The study was funded in part by a NICHD Training Grant (HD07289) to Dr. D. W. Fulker. Preparation of the paper was facilitated by grant RR-07013-20 awarded to the University of Colorado by the Biomedical Research Support Grant Program, Division of Research Resources, National Institutes of Health. DPostdoctoral fellow, Institute for Behavioral Genetics, University of Colorado. Received Ph.D. in psychology from University of Virginia. Current research interests are intelligence and prosocial and antisocial behaviors from a developmental behavior genetics perspective. "Research Scientist, Center for Community Research and Action, New York University. Received Ph.D. in psychology from Michigan State University. Current research interests are developmental pathways to problem behaviors of youth in high-risk communities. 3Ph.D. candidate at University of Virginia. Current research interests are volunteer interventions with adolescents at risk for delinquency. 'Ph.D. candidate at University of Virginia. Current research interests are clinical applications and intervention with adolescents and families. STo whom reprint requests should be addressed at Institute for Behavioral Genetics, University of Colorado, Boulder, CO 80309-0447. 233 0047-2891/88/0600-0233506.00/0 9 1988 Plenum Publishing Corporation

234

DiLalla et al.

INTRODUCTION Juvenile delinquency has become an alarming concern to society (Empey, 1982), affecting youths from all walks of life (Gold and Petronio, 1980). Over 80~ of American adolescents rated themselves as having engaged in at least one delinquent act (Gold and Petronio, 1980). In particular, aggressive delinquent behaviors are the most harmful to society and are increasing in prevalence. For example, Shamsi (1981) reported a 300070 increase in arrests of 7- to 17-year-olds for violent crimes, an increase almost double that for adults. A developmental perspective can be especially useful in studying the problem of delinquency and aggression by focusing on age-specific issues (Sroufe and Rutter, 1984). When assessing and intervening with adolescents, it is necessary to take into account such influences as the significance of the peer group and the adolescents' budding independence, and the changing roles of these factors within the family. Family variables that may contribute to delinquency have been the focus of interest in some studies (e.g., Glueck and Glueck, 1968, 1972; McCord, 1978; Patterson, 1986). A number of variables have been identified, including parental quarreling, lax discipline, parental and sibling delinquency, and extremely lax, harsh, or inconsistent punishment (Achenbach, 1982; Bandura and Waiters, 1959; Eron et al., 1963; Glueck and Glueck, 1968; Martin, 1975; Rutter, 1981). However, these variables have been studied less frequently with respect to aggressive development. One group of research studies has supported the idea that aggressive behaviors are learned in the family (Patterson, 1982, 1986; Patterson et al., 1975). They have suggested that children may develop aggressive tendencies through instinctual mechanisms (e.g., infants may cry in an impelling manner when they want something); reinforcement (e.g., a child learns that by hitting her brother, she can make her brother stop teasing her); and modeling (e.g., children may imitate aggressive behaviors when they see peers or family members yelling and hitting) (Patterson, 1982). Inept discipline practices by the parents are hypothesized to interact with child coerciveness and to be causally related to later antisocial behavior (Patterson, 1986). For example, a parent intrudes aversively into the child's activities, the child responds aggressively, and the parent withdraws. Thus, the child learns that aggression causes the parent to leave him alone, and the parent is punished for attempting to control the child and learns to be more lax. Such a sequence represents "escape-avoidant conditioning" (Patterson, 1986, p. 436). This and other theories about the acquisition of aggressive behaviors center around the youth's perceptions of others and imitations of others' acts. One particularly helpful approach is social learning theory, which states that one's cognitive perceptions of the environment affect one's own personality and actions (Parke and Slaby, 1983; Ryckman, 1982). There is strong sup-

Aggression and Delinquency

235

port in social learning theory for the notion that imitation of successful aggressive acts influences development of later aggression. A number of laboratory experiments have demonstrated that people shown aggressive behaviors that are rewarded are more likely to imitate those behaviors (Bandura, 1969; Meyer, 1972; Mischel, 1981; Ryckman, 1982). This theory can also apply to delinquent behaviors. It is likely that youths surrounded by delinquent peers or deviant family members will imitate their behaviors. This is especially true if the model is someone who is important to or controls the rewards for the youth. Hence, at adolescence when peer status becomes increasingly important (Short, 1966), delinquent peers should provide a powerful modeling stimulus. These findings may have important consequences for the study of both delinquency and aggressive tendencies. Children may be imitating violent and aggressive behaviors that they witness in the home; thus, modeling may be a factor in the development of aggressive behaviors (Bandura and Waiters, 1959). When children misbehave, they may be disciplined violently. In turn, they see that their parent gets what he or she wants; that is, the noxious behaviors that the child was exhibiting is stopped. According to social learning theory, this provides the child with information about how to obtain reinforcement, and the child is then more likely to imitate the aggressiveness. People also are more likely to imitate those who are like them, who are admired, and who dispense reinforcers (Ryckman, 1982). Therefore, it is easy to understand why children might imitate their parents when their parents use violent punishments. However, the actual situations in which children imitate require further exploration (Parke and Slaby, 1983). In addition to studying aggressive and delinquent behaviors globally, it is useful to examine subtypes of delinquency. Although delinquent behaviors have been classified into several subtypes (Quay, 1972; Quay and Peterson, 1983), few studies have attempted to predict youths' delinquency ratings based on family and environmental variables. Certain factors may be differentially related to type of delinquency. This is important when studying possible precursors to delinquency as well as when determining whether certain interventions are effective. Different types of delinquency actually may arise from very separate causes and may respond to quite disparate treatments. Two youth behaviors were of interest in this study: delinquency and aggressiveness. The factors under consideration were (a) level of violence in parental discipline, (b) the youth's perception of peers or family members behaving in a deviant (i.e., delinquent or socially inappropriate) manner in the youth's environment, and (c) the amount of arguing and turmoil that exists within the youths' homes that may be imitated by the youths. It also was expected that the youths' levels of aggressiveness would be a function of family backgrounds and peers, and that the modeling behaviors of those around the youths would affect the youths' overall levels of aggressiveness.

236

DiLalla et aL

METHODS Subjects The subjects were juveniles from a semirural university community in the Southeast. They were part of an ongoing intervention project for juvenile delinquents. Involved in this project at the time of analysis were 18 youths with the following characteristics: mean age was 14.7 years (range from 12 to 17 years), 17 males, 12 of whom were white; the 1 female was white; 3 youths currently were not enrolled in school; mean parent education level was 12 years. The offenses for which youths were referred to this project ranged from trespassing to assault, with probation violation and petty larceny the most common.

Procedure Youths were referred to this project by either an intake counselor or judge; all youths had been charged for an offense other than a status offense. The youths and their parents were told that they could either join the project or proceed with processing as usual. All youths except 1 chose to participate in the project. The youths and parents were interviewed within 48 hours after agreeing to participate in the program. Seventeen of the parents interviewed were mothers; 1 was a father. The interviews were conducted by trained undergraduate volunteers at the University of Virginia. The interview was semistructured; the interviewer talked individually with the youth and the parent about the items in the interview and audio-recorded their answers; answers were coded following the interview according to predetermined scales. In addition, the youth and parent were administered several checklists. All questions were answered with respect to the events of the previous six weeks.

Measures Life Domain Measures Eight of the questions used in this study were taken directly from measures developed elsewhere on a similar project (Mitchell, 1980). These questions were combined with three questions designed specifically for this study. Together, these items were meant to tap three major areas: punishment, opinions of delinquency, and family turmoil.

237

Aggression and Delinquency

One question was pertinent to the area of punishment. Punishment referred to any method the parents or disciplinarians employed using the presentation of negative stimuli or the removal of positive stimuli to try to alter the behavior of the youths. The item had a 7-point response scale, from being so lax that there was no punishment to using moderate forms of punishment (such as grounding the youth) to using highly punitive and aggressive forms of punishment. Second, the youths' opinions about others in the environment who engaged in illegal or delinquent behaviors were assessed two ways. First, the respondents rated the youths' opinions about the illegal or delinquent behaviors of peers on a scale ranging from highly admire to highly disapprove. Second, questions regarding acting up in school and misbehaving with peers were included because the youths' behaviors were considered indicative of their opinions on the appropriateness of others' behaviors. Three items covered the "attitudes" area. Third, family turmoil was quantified by seven questions concerning parental disagreement over child discipline and child rearing, frequency of discipline in the past six weeks, and frequency of arguing in the home. These questions were expected to tap negative behaviors in the family that might be related to aggressiveness. A rational-empirical approach to measure construction was used (Jackson, 1970). The internal consistency of the scales was tested using Cronbach's alpha (Cronbach, 1970) and is presented in Table I. Due to the high intercorrelation of the Turmoil and Opinion scales, the two scales were combined into an Environmental Turmoil scale. However, the Opinion subset scale was maintained as a separate scale as well because certain analyses were only concerned with youths' opinions of delinquent peers. Furthermore, since this scale was internally reliable, it could be used as a single entity. The aggressiveness of punishment question was maintained as a separate item because it did not fit on either of the scales, but was important conceptually. Interrater reliabilities were assessed for the items on the Environmental Turmoil and Opinion scales and the Punishment question. The percentage of raters who had 80~ or better agreeement per item ranged from 73 to 100070 (mean of 92070). The average point difference for each item ranged from .06 to 1.27 (mean of .52), suggesting that the raters were indeed coding answers similarly.

Aggressive Behavior The Revised Behavior Problem Checklist (RBPC; Quay and Peterson, 1983) was used to measure parents' perceptions of aggressive behavior exhibited by the youths. The checklist's 80 items had a 3-point response set,

DiLalla

238

a

aL

Table 1. Reliabilities for Life Domain Scales Corrected item,to-total correlation

Items Turmoil scalC Parents and youth argue Siblings and youth argue Parents disagree about discipline Argue-where youth is going Argue-- chores Argue-youth's friendsFrequency of discipline

.78 - .04 b

.72 .65 .69 .68 .00~

Opinion scalec Youth gets in trouble with teachers Youth and peers get in trouble Opinion of peers getting in trouble

.77 .67 .70

Family scale~ Parents do anything illegal? Have they gone to court? Sibs do anything illegal? Have they gone to court?

.41 .26 .42 .55

*Cronbach's alpha = .78. bltems were maintained, even with low item-total correlation, because the alpha was still high and the items were of interest. "Cronbach's alpha = .83. dCronbach's alpha = .62. f r o m 0 (not a problem at all) to 2 (frequently a problem). Only t w o o f the checklist's six problem behavior categories were o f interest in this study: conduct disorder (CD), or aggressive, interpersonally alienated, acting-out behavior; a n d socialized aggression (SA), or acting-out, authority-rejecting behavior but with strong social bonds. Q u a y and Peterson (1983) have based these distinctions o n those made between undersocialized and socialized conduct disorders according to D S M - I I I (American Psychiatric Association, 1980). Using the a b o v e procedure f o r measure development, two scales o f 11 items each emerged (see Table II). Because the scale intercorrelation was fairly high (see Table III), the two scales were c o m b i n e d as an overall Aggressiveness scale, with an internal consistency coefficient o f .93.

Referral Data C o u r t records were examined after the first g r o u p o f youths ( N = 13) h a d finished intervention. Types o f referral offenses, as well as n u m b e r o f offenses f o r the year previous to intervention, were noted. The d a t a were collected by a graduate student blind to the hypotheses being tested. T h e referral charges were rated in terms o f level o f aggressiveness, f r o m 1 (nondes-

Aggression and Delinquency

239 Table II. Reliabilities for R B P C Corrected total-to-item correlation*

Items

CD

Disruptive, annoys others Fights Negative Impertinent Irritable, easily angered Persistent and nags Bullies, threatens Picks at other kids to get attention W a n t s everything now W o n ' t do as told Blames others Stays out late T r u a n t from school with others Has "bad" companions A d m i t s disrespect for moral values Rejects school activities along with group Cheats Seeks c o m p a n y o f older companions Lies to protect friends Uses alcohol with others Admires rougher peers Admires those who operate outside the law

SA

Total Aggression

.36 .42 .40 .73 .19 .47 .53 .45 .49 .76 .36

.48 .69 .78 .89 .87 .68 .74 .64 .93 .89 .87 .20 .11 .46 .83 .17 .60 .36 .48 .31 .44 .36

.55 .71 .68 .88 .85 .77 .58 .63 .85 .86 .92

*Cronbach's alphas: CD, .94; SA, .81; Total Aggression, .92.

tructive offenses against property) t o 3 (destructive offenses against property) (offenses that involved aggressiveness toward another individual).

to 5

In sum,

the questions

(I) modeling behaviors

used in this study covered

and the environment,

the following

areas:

via questions from the Environ-

Table IlL lnterscale Correlations Punishment Youth Punishment Youth Parent Opinion Youth Parent Turmoil Youth Parent RBPC Referral

-

Parent .82 b -

*Cannot be computed. bp < .05; N ranged from 4 to 13.

Opinion Youth - .80 b .47 -

Turmoil

Parent

Youth

Parent

RBPC

.00 .19

- .68 .55

- .62 .19

.00 .29

- .65 b - .66 b

.26 -

.88 .51

.87 b .62

* - .20

.97 b .76 b --

* - .04 .25

-

.96 b .90 b .72 -

Referral

240

DiLalla et

al.

mental Turmoil, Opinion, and Punishment scales; (2) parents' perceptions of youths' aggressive behaviors via the RBPC, which provided the three variables of SA, CD, and overall aggressiveness; and (3) level of aggression using court referral data. The two hypotheses were that (1) violence in the home and high opinions of deviant peers would be related to high socialized aggressiveness ratings of the youths, and (2) environmental turmoil and aggressive punishments would be related to aggressiveness of the youths.

RESULTS Relationship o f Peers and Punishment with SA Rating

The first hypothesis, based on concepts of modeling and delinquency, was that the more impressed youths were with deviant peers and the more aggressively youths were punished, the more likely they would be to have higher ratings of delinquency with peers. The independent variables were the scores on the Opinion scale and the Punishment question; the dependent variable was the youths' scores on the Socialized Aggressiveness scale of the RBPC. Only the youths' ratings were used in this analyses for the independent variables because the parent-youth correlations on the Opinion scale were low and the youths' perceptions were of primary importance. A univariate regression analysis revealed that the independent variables were significantly related to SA scores. The Opinion variable contributed significantly to the regression: youths with high Opinion scores (e.g., thinks highly of deviant friends, gets in trouble with peers) also had high scores on SA. Punishment was not a significant predictor (see Table IV). Thus, the results showed that youths' opinions of their deviant peers were statistically predictive of the magnitude of youths' SA scores. Contrary to what had been expected, however, aggressiveness of Punishment in the home had no effect on the youths' SA scores. Table IV. RegressionAnalysis:Predictionof Total

SA Score Multivariate Adjusted R 2 .529 Covariate Opinion

Punishment

df

F

2,14

9.99

Univariate /3 t .718 3.975 - .127 - .706

Significance .002 Significance .001 .492

Aggression and Delinquency

241

Relationship of Punishment and Environmental Turmoil with Aggressiveness

The second hypothesis, also based on social learning theory, stated that youths who had higher negative ratings on the Environmental Turmoil scale and who had been punished more aggressively would have higher scores on the overall Aggressiveness scale f r o m the RBPC and the court referral data. The parent and youth responses for both o f the independent variables were highly correlated; therefore, the parent and youth scores for each family were averaged. In this way, each family had one Environmental Turmoil and one Punishment score. As can be seen in Table V, a multivariate multiple regression revealed that there was a significant effect due to regression. The variables predicting the R B P C aggressiveness scale accounted for 51% o f the variance. In partial support o f the hypothesis, the Environmental Turmoil scale was a significant statistical predictor o f reported aggressiveness on the RBPC. However, Punishment was not a predictor. The second dependent variable was the aggressiveness of the court referral data; 46~ of the variance was accounted for by the variables. U p o n examination, only Punishment was found to be a significant statistical predictor o f aggressiveness. The slope of the regression line was negative, showing that youths punished most aggressively (e.g., yelled at or hit) were referred for offenses that were less aggressive, and youths punished the least aggressively (e.g., not at all or just talked to) were referred for more aggressive offenses. Ratings on the Environmental Turmoil variable did not predict aggressiveness of referral offense at all.

Table V. Regression Analysis of Predictors of Aggressiveness

Multivariate Pillai's criterion, F(4,28) = 6.49 p < .001 Univariate F tests Dependent variable RBPC Referral

AdjustedR2 df F .443 2,14 7.355 .386 2,14 6.019 Univariate regression analysis

Dependent variable

Covariate

RBPC

Turmoil Punishment Turmoil Punishment

Referral

/~ .682 .285 - .044 - .682

t Value 3.639 1.519 - .224 - 3.469

Significance .007 .013 Significance .003 .151 .826 .004

242

DiLalla e t al.

Because the results for the Punishment variable were the inverse of what had been expected, further exploration into the nature of the Punishment variable was necessary. Punishment was trichotomized into the following categories: never punished or simply talks to youth (Level 1); withdraws positives, grounds, or adds negatives (Level 2); yells or spanks (Level 3). A regression analysis using this variable and Environmental Turmoil as predictors for aggressiveness o f referral revealed that Level 1 Punishment was the only significant predictor (t = 3.91, p < .002) and predicted the most aggressive offenses. The correlation between age of youth and Punishment score was examined in order to assess whether punishments differed simply as a function of age. The correlation was significant when the youths' Pumshment ratings were used (r = .51, p < .03); as age increased, punishments were reported as more aggressive. However, the correlation was nonsignificant when the parent ratings were used. Correlations between age and all other variables in this study were also examined, and no others were significant. In sum, high environmental turmoil was predictive of high ratings of the youths on the overall aggressiveness scale of the RBPC. Also, aggressiveness of punishment was found to predict the aggressiveness of the referring offense. However, this relationship was in the opposite direction of what had been hypothesized; youths who had been punished aggressively were likely to have been referred for less aggressive offenses, while those who were not punished at all or were punished very little were likely to have been brought in for the most aggressive offenses.

DISCUSSION Four possible limitations of this study should be considered: most of the variables are based only on self-report measures; the directions of effects or the impact of unexamined variables cannot be discerned; and because of stability concerns inherent in any small sample, significant findings require replications. Finally, where expected results were not found, we must always be wary of accepting the null hypothesis (Cook and Campbell, 1979; Spector, 1981). However, although the sample size was small, the findings reached significance and accounted for a large proportion of the variance (approximately 5007o). With these caveats in mind, interesting findings can be presented. A number of consistent results were found that were highly significant. These are summarized below, and theoretical implications are discussed9 Parents' perceptions about their children's aggressiveness and the youths' most aggressive acts were shown to be independent and related to different

Aggression and Delinquency

243

variables. Parents' perceptions of aggressive problem behaviors of the youths were related to the youths' opinions of these friends and to high turmoil in the home. One aspect of aggressiveness of the youths' referring offenses was related to the behaviors of the parents in terms of punishment. These interesting relationships suggest that opinions influenced cognitive perceptions but not behaviors (youths' opinions of their friends influenced parents' perceptions of youths' behaviors but not aggressiveness of the youths' referring offenses), whereas actual behaviors influenced each other (punishment in the home was related to actual aggressiveness of referring offense). Contrary to expectation, aggressiveness of punishment in the home did not statistically predict how highly youths were rated on socialized aggression. There are three possible explanations for this. First, as noted above, SA was rated by the parents and reflected their perceptions of their children's behaviors. It may be that the parents' own behaviors in the home (the punishments they use) do not impact on their cognitive impressions of their children's behaviors. Also, the parents may not punish their children for the sorts of behaviors listed in the RBPC. Second, extremely aggressive punishments were never reported. Hitting was only reported by three families, suggesting that the families in this sample may have been reluctant to mention any punishments that were much more severe, especially to interviewers who represented both the court system and psychologists. Finally, the RBPC measure does not differentiate between CD and SA youths based on aggressiveness of their behavior. If punishment is only related to aggressive behaviors and not to levels of socialization, the the RBPC subtypes would not be predictable from forms of punishment used in the home. The results also showed that enviromental turmoil was related to high scores on the overall aggressiveness scale of the RBPC. However, aggressiveness of punishment was not related to overall aggressiveness, possibly for the same reasons outlined above. On the other hand, punishment was a good statistical predictor for aggressiveness of the referring offense, although the relationship was in the opposite direction of what had been predicted. Initially this may appear confusing, especially as it does not support the idea that children imitate their parents' behavior. However, this result parallels literature showing that children who are not at all disciplined tend to be more unruly, delinquent, and aggressive (Bandura and Waiters, 1959; Glueck and Glueck, 1968; Patterson et al., 1975). In fact, it was found that only the lax punishment response was related to aggressiveness of offense. The results from this study can be explored in a number of ways by placing them within the broader theoretical framework of developmental and social learning theory. From a developmental perspective, these results can partially be explained by the peer group's increasing importance during adolescence (Short, 1966). At this time, youths' opinions of peers would be expected

244

DiLalla et aL

to affect the youths' behaviors. Youths who think highly of their delinquent friends might behave in a manner that would cause their parents to rate them as highly socialized aggressive. It is important to recognize that the youths' opinions of their friends were related only to their parents' perceptions of their behaviors. The opinions were not related to the aggressiveness of the referring offense; additionally, there was no relationship between opinion~ and youths' self-reports of their recent delinquent behaviors (DiLalla, 1985). Therefore, it is possible that the relationship is a function of the parents' frustrations with the general friendship situations, and that the parents did not distinguish between the behaviors of their children and of their children's friends. Social learning theory adds to the explanation of why youths who lived in a turmoiled environment were rated by their parents as having aggressive problem behaviors. Growing up in an aggressive environment is a long-term process that may affect current and later behavior. Youths who were surrounded by parents and siblings who argued frequently and who fought each other in order to get what they wanted may have learned that arguing and acting aggressively were effective ways of getting rewards. This study suggests that different processes influence perceptions of aggressive problem behaviors and actual aggressive delinquent offenses. The former are influenced by perceptions of other aspects of the youths' lives, while the latter are related to behaviors in the home that affect the youths. Although several variables have been discovered that may be useful in discussing aggression and delinquency, much research is still needed to complete the picture of the development of aggression and delinquency. If the causes of these behaviors can be found, then interventions should be more likely to succeed. The direction of causality in the relationship between discipline practices and youth behavior is as yet unknown. For example, discipline practice may be an important variable to attempt to modify through intervention; however, the child's temperament may cause the parents to react to the child more or less aggressively as well (Patterson, 1986). Finally, parents in turmoiled households may tend to have children with certain aggressive or coercive temperaments. This study suggests the importance of examining aggressiveness in adolescence developmentally and across environments.

ACKNOWLEDGMENTS For their moral and editorial help, the authors wish to thank David DiLalla, Julie Criss, and Jon Heeren. We gratefully acknowledge the undergraduates who helped with data collection and intervention, and the court

Aggression and Delinquency

245

a n d the families w h o c o o p e r a t e d with this project. A n a n o n y m o u s reviewer was especially h e l p f u l o n a n earlier draft.

REFERENCES Achenbach, T. M. (1982). Developmental Psychopathology. John Wiley & Sons, New York. American Psychiatric Association (1980). Diagnostic and Statistical Manual of Mental Disorders, 3rd ed. American Psychiatric Association, Washington, DC. Bandura, A. (1969). Principles of Behavior Modification. Holt, Rinehart, & Winston, Inc., New York. Bandura, A., and Waiters, R. H. (1959). Adolescent Aggression. Ronaid Press, Co., New York. Cook, T. D., and Campbell, D. T. (1979). QuasiExperimentation. Houghton Mifflin, New York. Cronbach, L. J, (1970). Essentials of Psychological Testing, Harper & Row, New York. DiLalla, L. F. (1985). Juvenile delinquency: A study of "turmoil" and aggression. Unpublished masters thesis, University of Virginia. Empey, L. T. (1982).American Delinquency: Its Meaning and Construction. DorseyPress, Homewood, IL. Eron, L. D., Waider, L. O., Toigo, R., and Lefkowitz, M. M. (1963). Social class, parental punishment for aggression, and child aggression. Child Develop. 34: 849-867. Gold, M., and Petronio, R. J. (1980). Delinquent behavior in adolescence. Adelson, J. (ed.), Handbook of Adolescent Psychology. John Wiley & Sons, New York. Glueck, S., and Glueck, E. (1968). UnravelingJuvenile Delinquency. Harvard University Press, Cambridge, MA. Glueck, S., and Glueck, E. (1972). Identification of Ire-delinquents: Validation Studies and Some Suggested Uses of Glueck Table. Intercontinental Medical Book Corporation, New York. Jackson) D. N. (1970). A sequential system for personality scale development. In Spielberger, C. (ed.), Current Topics in Clinical and Community Psychology, Vol. 2 Academic Press, New York. Martin, B. (1975). Parent-child relations. In Horowitz, F. D. (ed.), Review of Child Development Research, Vol. 4. University of Chicago Press. McCord, J. (1978). A thirty-year follow-up of treatment effects. Am. Psychol. 33: 284-289. Meyer, T. P. (1972). Effects of viewingjustified and unjustified real film violence on aggressive behavior. J. Personal. Social Psychol. 23: 21-29. Mischel, W. (1981). Introduction to Personality. Holt, Rinehart, & Winston, Inc., New York. Mitchell, C. M. (1980). Nonprofessionaisworking with delinquent youth: An experimentalcomparison of university, community collegeand community nonprofessionais. Unpublished doctoral dissertation, Michigan State University, MI. Parke, R. D., and Slaby, R. G. (1983). The development of aggression. In Mussen, P. (ed.), Handbook of Child Psychology, Vol. 4. John Wiley & Sons, New York. Patterson, G. R. (1982). Coercive Family Processes. Castilia Press, Eugene, OR. Patterson, G. R. (1986). Performance models for antisocial boys. Am. Psychol. 41(4): 432-444. Patterson, G. R., Reid, J. B., Jones, R. R., and Conger, R. E. (1975). A Social Learning Approach to Family Intervention. Vol. I: Families with Aggressive Children. Castalia Publishing Co., Eugene, OR. Quay, H. C. (1972). Patterns of aggression, withdrawal, and immaturity. In Quay, H. C., and Werry, J. S. (eds.), Psychopathological Disorders of Childhood. John Wiley & Sons, New York. Quay, H. C., and Peterson, D. R. (1983). Interim manual for the Revised Behavior Problem Checklist, 1st ed. Unpublished. Rutter, M. (1981). Stress, coping and development: Some issues and some questions. J. Clin. Psychol. Psychiatr. 22(4): 323-356.

246

DiLalla et al.

Ryckman, R. M. (1982). Theories of Personality, 2nd ed. Brooks-Cole Publishing Co., Monterey, CA. Shamsi, S, J. (1981). Antisocial adolescents: Our treatments do not w o r k - W h e r e do we go from here.'/Can. J. Psychiatry. 26: 357-364. Short, J. F., Jr. (1966). Juvenile delinquency: The sociocultural context. In Hoffman, L. W., and Hoffman, M. L. (eds.), Review of Child Development Research, Vol. 2. Russell Sage Foundation, New York. Spector, P. E. (1981). Multivariate data analysis for outcome studies. Am. J. Commun. Psychol. 9: 45-53. Sroufe, L. A., and Rutter, M. (1984). The domain of developmental psychopathology. Child Develop. 55: 17-29.