Predictors of Youth Violence by pjv36417

VIEWS: 144 PAGES: 12

More Info
									U.S. Department of Justice
Office of Justice Programs
Office of Juvenile Justice and Delinquency Prevention

    John J. Wilson, Acting Administrator                                                                               April 2000

Predictors of Youth                                                                             From the Administrator
Violence                                                                                        If we could confidently predict which
                                                                                                youth would be prone to commit
                                                                                                violent acts and at which stage in
                                                                                                their development such delinquency
                                                                                                was most likely to erupt, it would
                                                                                                significantly strengthen our efforts
J. David Hawkins, Todd I. Herrenkohl, David P. Farrington,                                      to prevent juvenile violence.
Devon Brewer, Richard F. Catalano, Tracy W. Harachi, and                                        Accordingly, the Office of Juvenile
Lynn Cothern                                                                                    Justice and Delinquency Prevention’s
                                                                                                (OJJDP’s) Study Group on Serious
Identifying and addressing the predictors         analysis procedures. The 66 studies exam-     and Violent Juvenile Offenders
of youth violence at appropriate points in        ined here were drawn from Lipsey and          devoted 2 years to analyzing the
youth development is important for pre-           Derzon’s bibliography (1998) and supple-      research on risk and protective
vention. Unfortunately, there have been           mented by research reports provided by        factors for serious and violent
few high-quality longitudinal studies of          OJJDP Study Group members and analy-          juvenile offending, including predic-
the predictors of youth violence. The Of-         ses of the Seattle Social Development         tors of juvenile violence derived
fice of Juvenile Justice and Delinquency          Project longitudinal data set. The studies    from the findings of long-term
Prevention’s (OJJDP’s) Study Group on             selected for this review met the following    studies.
Serious and Violent Juvenile Offenders            six criteria:
(Study Group) brought 22 researchers                                                            This Bulletin describes a number
together for 2 years to analyze current           x Subjects were juveniles living in their     of such risk and protective factors,
                                                    community (i.e., they were not incarcer-    including individual, family, school,
research on risk and protective factors
and the development of serious and vio-             ated) when they were first assessed.        peer-related, community/neighbor-
lent juvenile offending careers.                  x Subjects were not chosen for having         hood, and situational factors.
                                                    committed prior criminal or violent         Although we need additional
Together, data from the long-term studies           offenses.
that have identified predictors of youth                                                        research on juvenile violence, the
violence can help determine violence pre-         x Studies measured interpersonal physi-       information this Bulletin provides
vention policy and practice. This Bulletin          cal violence or acts resulting in physi-    will enhance our understanding of
describes the strength and duration of              cal injury or threat of physical injury     the predictors of youth violence. I
changeable risk and protective factors for          to another person, excluding suicidal       would also call your attention to
youth violence at points in youth develop-          behavior.                                   the Study Group Report and to the
ment when they appear most salient. These                                                       Bulletin summarizing it, both of which
                                                  x Studies identified a modifiable indicator
predictors are potential targets for preven-                                                    may be obtained from OJJDP’s
                                                    of a meaningful predictor or risk factor.
tion and intervention. If risk factors can be                                                   Juvenile Justice Clearinghouse.
                                                    Studies of interactions between mul-
decreased and protective factors enhanced           tiple risk factors were excluded, as were   John J. Wilson
by preventive action, then the likelihood           discussions of race and gender, as pre-     Acting Administrator
of violence should be reduced.                      dictors of violence.
                                                  x The study design was longitudinal,
Study Sample                                        with results based on prospective or
                                                    retrospective data so that exposure
The quantitative results of a large number
                                                    to risk factors preceded violence.
of studies were synthesized using meta-
x Individual subjects served as the unit       x Family factors:                          Low resting heart rate. This predictor is
  of analysis for both independent and           y Parental criminality.                  thought to indicate a fearless tempera-
  dependent variables.                                                                    ment or underarousal, which may predis-
                                                 y Child maltreatment.                    pose an individual to aggression and vio-
                                                 y Poor family management practices.      lence (Raine and Jones, 1987). Research
Methodology                                                                               indicates that a low resting pulse rate
A statistical analysis was performed to          y Low levels of parental involvement.
                                                                                          is a weak predictor of violent crime
determine the strength of the association        y Poor family bonding and family         (Farrington, 1998; Wadsworth, 1976).
between particular risk factors and the            conflict.
violence incurred. To account for the fact                                                The evidence currently does not war-
                                                 y Parental attitudes favorable to        rant using either of these predictors—
that each study used different methods,            substance use and violence.
this relationship was expressed as a cor-                                                 pregnancy and delivery complications
relation coefficient, which was arrived at       y Parent-child separation.               or low resting heart rate—to identify
using standard meta-analytical proce-                                                     youth at risk for violent behavior. More
                                               x School factors:
dures (Rosenthal, 1991). The findings                                                     research is needed on these factors and
from two or more studies were summa-             y Academic failure.                      their possible effects on violence.
rized as a weighted mean correlation,            y Low bonding to school.
which gives more weight to studies with          y Truancy and dropping out of            Individual Psychological
large samples than to studies with small           school.                                Factors
samples.                                                                                  Internalizing disorders (nervousness/
                                                 y Frequent school transitions.
The strength of the association between                                                   withdrawal, worrying, and anxiety).
                                               x Peer-related factors:                    This category of psychological charac-
a risk factor and subsequent violence
can also be expressed as an odds ratio           y Delinquent siblings.                   teristics has a slight negative correla-
(the odds of violence in the group with a                                                 tion with (Mitchell and Rosa, 1979),
                                                 y Delinquent peers.
particular risk factor divided by the odds                                                or is unrelated to, later violence
of violence in the group without that risk       y Gang membership.                       (Farrington, 1989).
factor). An odds ratio expresses the de-       x Community and neighborhood               Hyperactivity, concentration problems,
gree of increased risk for violence associ-      factors:                                 restlessness, and risk taking. Evidence
ated with the presence of a risk factor in       y Poverty.                               from studies in this meta-analysis
a population. For example, an odds ratio                                                  consistently suggests a correlation be-
of 2 indicates a doubling of risk. This Bul-     y Community disorganization.
                                                                                          tween these problems and later violent
letin provides odds ratios for predictors        y Availability of drugs and firearms.    behavior.
when they were given or could be com-
piled from the information provided in           y Neighborhood adults involved in        In a longitudinal study in Sweden, 15
                                                   crime.                                 percent of boys with both restlessness
a study.
                                                 y Exposure to violence and racial        and concentration difficulties at age 13
                                                   prejudice.                             were arrested for violence by age 26.
Results                                                                                   Boys with restlessness and concentra-
Predictors are arranged in five domains:       Individual Medical and                     tion difficulties were five times more
individual, family, school, peer-related,      Physical Factors                           likely to be arrested for violence than
and community and neighborhood fac-                                                       boys without these characteristics
                                               Pregnancy and delivery complications.
tors. The following malleable predictors                                                  (Klinteberg et al., 1993).
                                               Prenatal and delivery trauma are some-
of violence are discussed in more detail       what predictive of later violence, al-     In another study, Farrington (1989) found
below.                                         though findings vary with the research     that teacher ratings of male children’s con-
x Individual factors:                          methods used.                              centration problems and restlessness—
                                                                                          including difficulty sitting still, the
   y Pregnancy and delivery complications.     Kandel and Mednick (1991) found that
                                                                                          tendency to fidget, and frequent
                                               80 percent of violent offenders scored
   y Low resting heart rate.                                                              talkativeness—predicted later violence.
                                               high in delivery complications, compared
   y Internalizing disorders.                                                             Concentration problems also predicted
                                               with 30 percent of property offenders
                                                                                          academic difficulties, which predict later
   y Hyperactivity, concentration prob-        and 47 percent of nonoffenders. How-
                                                                                          violence. Multivariate models are needed
     lems, restlessness, and risk taking.      ever, other studies have not found an
                                                                                          to understand the pathways leading to
                                               association between pregnancy and
   y Aggressiveness.                           delivery complications and violence
                                                                                          violent behavior.
   y Early initiation of violent behavior.     (Denno, 1990; Farrington, 1997). Mednick   Aggressiveness. Aggressive behavior
   y Involvement in other forms of anti-       and Kandel found in an earlier study       measured from ages 6 to 13 consistently
     social behavior.                          (1988) that a stable home environment      predicts later violence among males.
                                               served as a protective factor against      Many researchers have noted the conti-
   y Beliefs and attitudes favorable to        prenatal trauma.                           nuity in antisocial behavior from early
     deviant or antisocial behavior.                                                      aggression to violent crime (Loeber,

                                                                                             to commit violent crimes than those with
                                                                                             noncriminal parents.
                                                                                             In contrast, Moffitt (1987) found that adults
                                                                                             (ages 29–52) with criminal parents were
                                                                                             not much more likely to be arrested for a
                                                                                             violent offense than those with noncriminal
                                                                                             parents. Further research is necessary to
                                                                                             understand the contribution of parental
                                                                                             criminality to child behavior.
                                                                                             The relationship between parental alco-
                                                                                             holism and mental illness and children’s
                                                                                             violent behavior has been examined.
                                                                                             McCord (1979) did not find a link be-
                                                                                             tween fathers’ alcoholism and criminal
                                                                                             conduct and their sons’ later violence.
                                                                                             In a study of male adoptees, Moffitt
                                                                                             (1987) found a small and inconsistent
                                                                                             relationship between parental mental
                                                                                             illness and violence in children.
1990, 1996; Loeber and Hay, 1996;              16 not adjudicated delinquent for a violent
Olweus, 1979). A study in Orebro,              crime as juveniles.                           Child maltreatment. Studies have exam-
Sweden, found that two-thirds of boys                                                        ined three forms of child maltreatment:
                                               Involvement in other forms of antiso-
with high teacher-rated aggression                                                           physical abuse, sexual abuse, and ne-
scores at ages 10 and 13 had criminal          cial behavior. Involvement in antisocial
                                                                                             glect. Evidence suggests that children
                                               behaviors, including stealing and de-
records for violent offenses by age 26.                                                      who have been physically abused or
They were more than six times more             struction of property (Mitchell and
                                                                                             neglected are more likely than others
                                               Rosa, 1979); self-reported delinquency,
likely than boys who were not rated                                                          to commit violent crimes later in life
aggressive to be violent offenders             smoking, and early sexual intercourse
                                                                                             (Widom, 1989; Zingraff et al., 1993;
                                               (Farrington, 1989); and drug selling
(Stattin and Magnusson, 1989).                                                               Smith and Thornberry, 1995).
                                               (Maguin et al., 1995), is associated
In a sample of African American boys in        with a greater risk of violence among         Poor family management practices. Fam-
the Woodlawn area of Chicago, IL, nearly       males. Robins (1966) found a similar          ily management practices such as failure
half of the 6-year-old boys who had been       pattern among male psychiatric patients       to set clear expectations for children’s
rated aggressive by teachers were ar-          but did not find similar patterns for         behavior, poor monitoring and supervi-
rested for violent crimes by age 33, com-      females.                                      sion, and severe and inconsistent disci-
pared with one-third of their nonaggres-                                                     pline consistently predict later delin-
                                               Beliefs and attitudes favorable to devi-
sive counterparts (McCord and Ensminger,                                                     quency and substance abuse (Capaldi and
1995). This relationship also held for males   ant or antisocial behavior. Dishonesty,
                                                                                             Patterson, 1996; Hawkins, Arthur, and
                                               antisocial beliefs and attitudes, atti-
in hyperactive samples (Loney, Kramer,                                                       Catalano, 1995). In a sample followed up
and Milich, 1983).                             tudes favorable to violence, and hostil-
                                                                                             on after 20 years, the McCords found that
                                               ity toward police have been found to
                                                                                             parents’ poor supervision and aggressive
Research results for females are less con-     predict later violence among males.
                                                                                             discipline predicted their children’s con-
sistent. McCord and Ensminger (1995)           Relationships between these predictors
                                                                                             victions for person crimes well into their
found similar results for males and fe-        and violence are less consistent for
                                                                                             forties (McCord, McCord, and Zola, 1959;
males; however, Stattin and Magnusson          females (Williams, 1994). Prevention
                                                                                             McCord, 1979).
(1989) did not find a relationship be-         programs that help youth develop
tween early female aggression and later        positive beliefs and standards so that        Wells and Rankin (1988) found that boys
violent offenses.                              they can reject violence, cheating, and       with very strict parents reported the most
                                               rule breaking may reduce the risk for         violence. Boys with very permissive par-
Early initiation of violent behavior.
                                               violence.                                     ents reported the second highest level of
Research has shown that early onset of
                                                                                             violence. Boys with parents who were nei-
violence and delinquency is associated
                                               Family Factors                                ther too strict nor too lax reported the
with more serious and chronic violence
                                               Parental criminality. Baker and Mednick       least violence. Also, boys whose parents
(Farrington, 1991; Piper, 1985; Thornberry,
                                                                                             punished them inconsistently, sometimes
Huizinga, and Loeber, 1995; Tolan and          (1984) found that men ages 18–23 with
                                               criminal fathers were 3.8 times more          punishing and sometimes ignoring the
Thomas, 1995). Farrington (1995) found
                                                                                             same behavior, were more likely to com-
that one-half of boys adjudicated delin-       likely to have committed violent criminal
                                               acts than those with noncriminal fathers.     mit an offense against other persons than
quent for a violent offense between age 10
                                                                                             boys whose parents punished them more
and age 16 were convicted of a violent         Farrington (1989) also found that boys
                                               who had a parent arrested before their        consistently. Parental punitiveness or
crime by age 24, compared with only 8 per-
                                                                                             harshness in discipline also predicted
cent of juveniles between age 10 and age       10th birthday were 2.2 times more likely
                                                                                             later violence.

Farrington (1989) found that poor child-      but little research has examined the im-      of violence in both men and women
rearing; an authoritarian parenting style;    pact of parental attitudes to violence on     (McCord and Ensminger, 1995). However,
poor parental supervision; harsh parental     children’s behavior. One study showed         many other factors that also predict vio-
discipline; a cruel, passive, or neglectful   that children who at age 10 had parents       lence can contribute to parent-child
parenting attitude; and parental disagree-    who were tolerant of violent behavior were    separations. Multivariate studies are
ment about childrearing each predicted        more likely to report violent behavior by     needed to understand the interactions
later violence. Maguin and colleagues         age 18 (Maguin et al., 1995).                 among these factors.
(1995) found that poor family manage-
ment practices when boys were ages            Residential mobility. Little research has
                                              focused on the effect of a family’s mobil-    School Factors
14–16 predicted self-reported violence
                                              ity on youth violence. Maguin and col-        Various aspects of school-related ex-
by age 18, although poor family manage-
                                              leagues (1995) found that the number          periences, such as low educational
ment practices when boys were age 10
                                              of changes in residence in the past year,     achievement, low interest in education,
did not predict violence at age 18. An
                                              assessed when boys were age 16, pre-          dropping out of school, truancy, and
analysis of a subsample of the Seattle
                                              dicted self-reported violent behavior by      poor-quality schools, have been hypoth-
Social Development Project data found
                                              18. Residential mobility assessed when        esized to contribute to criminal and vio-
that proactive family management prac-
                                              boys were age 14, however, did not sig-       lent behavior (Hawkins, Farrington, and
tices at age 14 reduced the likelihood of
                                              nificantly predict violence at age 18. This   Catalano, 1998).
self-reported violence at age 16 for Afri-
can American and Caucasian males and          discrepancy may indicate that residen-
                                                                                            Academic failure. Poor academic
                                              tial moves have short-term effects on
females (Williams, 1994).                                                                   achievement has consistently predicted
                                              behavior, but more research is needed
                                                                                            later delinquency (Maguin and Loeber,
Low levels of parental involvement.           to understand the relationship.
                                                                                            1996; Denno, 1990). Academic failure in
Strong parental involvement can function
                                              Parent-child separation. Evidence indi-       the elementary grades also increases risk
as a protective factor against violence.
                                              cates that disruptions of parent-child        for later violent behavior (Farrington,
Conversely, a lack of parental interaction
                                              relationships predict later violent behav-    1989; Maguin et al., 1995). The relation-
and involvement with children may in-
                                              ior in children. Parent-child separation      ship between poor academic achievement
crease children’s future risk for violence.
                                              before age 10 has been found to predict       and later violence has been found to be
Williams (1994) found that parent-child
                                              violence (Farrington, 1989; Wadsworth,        stronger for females than for males.
communication and involvement at age
14 predicted less self-reported violent       1978). Henry and colleagues (1996) found
                                                                                            Low bonding to school. Research gener-
                                              that having a single-parent family when
behavior at age 16. This relationship                                                       ally supports the hypothesis that bond-
was weaker for females than for males.        boys were age 13 predicted their convic-
                                                                                            ing to school is a protective factor
                                              tions for violence by age 18. An associa-
Similarly, Farrington (1989) found that                                                     against crime (Catalano and Hawkins,
sons whose fathers did not engage in lei-     tion also has been found between leaving
                                                                                            1996; Hirschi, 1969). Williams (1994)
                                              home at an early age and high levels
sure activities with them more often ex-                                                    found that school bonding was a stron-
hibited violent behavior as teenagers                                                       ger protective factor against violence in
and adults and were more likely to be                                                       African American students and in boys
convicted for a violent offense.                                                            and was less linked to violence in Cau-
                                                                                            casian students and in girls. Maguin and
Poor family bonding and conflict. Few
                                                                                            colleagues (1995) found that low com-
studies have looked specifically at the
                                                                                            mitment to school and low educational
relationship between family bonding and
                                                                                            aspirations at age 10 did not predict
violent behavior. Some research has
                                                                                            later violence, but at ages 14 and 16
shown a nonsignificant relationship be-
                                                                                            these factors increased the risk for vio-
tween poor family bonding and violence
                                                                                            lence. Other researchers have reported
(Williams, 1994; Elliott, 1994). Studies
                                                                                            that lack of school bonding was not a
investigating this link should distinguish
                                                                                            significant predictor of serious and vio-
between bonding to prosocial versus
                                                                                            lent offending (Elliott, 1994; Mitchell
antisocial or criminal family members
                                                                                            and Rosa, 1979).
(Foshee and Bauman, 1992).
                                                                                            Truancy and dropping out of school.
Exposure to high levels of marital and
                                                                                            Farrington (1989) found that youth with
family conflict also appears to increase
                                                                                            high truancy rates at ages 12–14 were
the risk of later violence (Farrington,
                                                                                            more likely to engage in violence as ado-
1989; McCord, 1979; Maguin et al., 1995;
                                                                                            lescents and adults; leaving school be-
Elliott, 1994).
                                                                                            fore the age of 15 also predicted later
Parental attitudes favorable to substance                                                   violence. Truancy and dropping out may
use and violence. Research indicates that                                                   be indicators of low school bonding, but
parental attitudes favorable to behaviors                                                   children also may miss school or leave
such as alcohol use predict use of alcohol                                                  school early for other reasons (Janosz
and drugs by youth (Peterson et al., 1994),                                                 et al., 1996).

Frequent school transitions. Maguin and
colleagues (1995) found that youth who
had changed schools often in the past year
at ages 14 and 16 were more violent at age
18 than those who had not. Conclusions
must be drawn carefully, however, because
school transitions can be related to other
factors that predict violence.
High delinquency rate school. Farrington
(1989) found that boys who at age 11 at-
tended schools with high delinquency
rates reported more violent behavior than
other youth.

Peer-Related Factors
Delinquent siblings. Farrington (1989)
found that having delinquent siblings by
age 10 predicted later convictions for
violence. Maguin and colleagues (1995)
                                             may contribute to crime and violence            Exposure to violence and racial prejudice.
found that the association between hav-
                                             (Brewer et al., 1995).                          Exposure to violence in the home and
ing delinquent siblings and being con-
                                                                                             elsewhere increases a child’s risk for in-
victed for violence was stronger when        Poverty. Being raised in poverty has been
                                                                                             volvement in violent behavior later in life
sibling delinquency occurred closer in       found to contribute to a greater likelihood     (Paschall, 1996). McCord and Ensminger
time to the violent youth’s offense and      of involvement in crime and violence
                                                                                             (1995) also found that African American
later in that youth’s development, indi-     (Sampson and Lauritsen, 1994). Self-            study participants who reported having
cating that antisocial siblings have a       reported felony assault and robbery have
                                                                                             experienced racial discrimination com-
stronger negative influence during their     been found to be twice as common among          mitted more violent acts.
sibling’s adolescence than earlier in        youth living in poverty as among middle-
the child’s development. Williams            class youth (Elliott, Huizinga, and Menard,
(1994) found that the influence of delin-    1989). Low family income predicted self-        Situational Factors
quent siblings was stronger on girls         reported teen violence and convictions          Situational factors are the circumstances
than on boys.                                for violent offenses in several studies         that surround a violent event and influ-
                                             (Farrington, 1989; Wikström, 1985; Hogh         ence the outcome of that event. These
Delinquent peers. Delinquent peers also
                                             and Wolf, 1983; Henry et al., 1996).            factors may be predictors of violent be-
may have a greater influence on later vio-
                                                                                             havior and may include the presence of a
lence during an individual’s adolescence     Community disorganization. Maguin and           weapon, consumption of alcohol or other
than they do earlier in development          colleagues (1995) examined community            drugs by the offender or victim, the be-
(Moffitt, 1993). Research has shown that     disorganization and low neighborhood            havior of bystanders, the motives of the
adolescents whose peers disapproved of       attachment as predictors of violence.           offender, the relationship of the offender
delinquent behavior were less likely to      Community disorganization (that is, the         to the victim, and the behavior of the
report having committed delinquent acts      presence of crime, drug-selling, gangs,         victim (Sampson and Lauritsen, 1994;
(Elliott, 1994), including sexual assaults   and poor housing) was a better predictor        Farrington and Loeber, 1999). However,
(Ageton, 1983).                              of violence than low attachment to a            the contribution of these factors is diffi-
                                             neighborhood.                                   cult to assess because data have not been
Gang membership. Battin and colleagues
(1998) showed that being a gang member       Availability of drugs and firearms. In          collected from other situations with simi-
contributes more to delinquency than         one study, a prevalence of drugs and            lar characteristics in which violence did
does having delinquent peers.                firearms in the community predicted             not occur. Longitudinal studies to investi-
                                             greater variety in violent behaviors at         gate these situational triggers are needed.
Community and                                age 18 (Maguin et al., 1995).
Neighborhood Factors                         Neighborhood adults involved in crime.
                                                                                             Multiple Predictors and
Community factors, including poverty,        Maguin and colleagues (1995) found that         Strength of Prediction
low neighborhood attachment and com-         children who knew many adult criminals          In the Seattle Social Development Project,
munity disorganization, the availability     were more likely to engage in violent be-       Herrenkohl and colleagues (in press) in-
of drugs and firearms, exposure to vio-      havior by age 18. More longitudinal studies     vestigated the power of diverse factors
lence and racial prejudice, laws and         investigating the influence of this factor on   seen at ages 10, 14, and 16 to predict
norms favorable to violence, and fre-        youth violence are needed.                      violent behavior by the age of 18. More
quent media portrayals of violence,

  Predictors of Violent or Serious Delinquency by Age Group: A Comparative Ranking1

  Researchers Mark W. Lipsey and James H. Derzon (1998) examined predictors of violent or serious delinquency in adolescence and
  early adulthood. Applying the procedures used for a meta-analysis, Lipsey and Derzon compiled information from published and
  unpublished research into a database that indexed the strength of the relationship between the predictor variable and the criterion
  variable in terms of effect sizes. Through a statistical analysis, the relative strength of different types of predictor variables was
  measured at different ages, and procedures were used to control for methodological differences between studies. The first goal
  was to determine which predictors seen at adolescence had the strongest empirical associations with subsequent violence or de-
  linquency. The second goal was to identify which of those associations were of sufficient magnitude to help identify at-risk juveniles
  to receive intervention.

  Results                                               involvement with antisocial peers—              Implications for Intervention
  The table on page 7 lists the predictors              have to do with interpersonal rela-             For an intervention to be effective, the tar-
  of violent or serious delinquency at                  tions. The same predictors, however,            geted risk factors must be amenable to
  ages 6–11 and ages 12–14 in the order                 are relatively weak for the 6–11 age            change. The strongest predictors of subse-
  of significance determined by the sta-                group.                                          quent violence for both age groups are
  tistical analysis and in groups based on          x Relatively fixed personal characteris-            relatively malleable factors. Because they
  estimated aggregated effect size.                   tics are the second- and third-rank               are cumulative, the second rank of varia-
                                                      predictors of subsequent violence                 bles for the 6–11 age group, the effects of
  The most interesting comparisons follow:
                                                      for the 6–11 age group. The ages                  antisocial parents and socioeconomic
  x The best predictors of violent or seri-           12–14 group has a heavier represen-               status, may not be very amenable to
    ous delinquency differ according to               tation of behavioral predictors of                change—and gender is not subject to
    age group. A juvenile offense at ages             subsequent violence (i.e., general                change. The predictors in the first, second,
    6–11 is the strongest predictor of                offenses, aggression, and school                  and third rank (except for male gender) for
    subsequent violent or serious delin-              performance).                                     juveniles ages 12–14 are malleable.
    quency even if the offense did not
                                                    x Broken homes and abusive parents                  Because many of the strongest predictors
    involve violence. For the 12–14 age
                                                      are among the poorest predictors of               of subsequent violence can be changed,
    group, a juvenile offense is the sec-
                                                      subsequent violence for both age                  they offer possible targets for successful
    ond most powerful predictor of future
                                                      groups.                                           intervention. This suggests that disrupting
    violence. Substance abuse is among
                                                                                                        early patterns of antisocial behavior and
    the best predictors of future violence          x The significance of antisocial peers              negative peer support is a promising strat-
    for children ages 6–11 but one of the             and substance abuse is reversed in                egy for the prevention of violence and
    poorest predictors for children ages              the two age groups. Whereas having                serious delinquency.
    12–14.                                            antisocial peers is a strong predictor
                                                      for the age 12–14 group, it is a weak             For more information about the meta-
  x The two strongest predictors of sub-
                                                      predictor for the age 6–11 group.                 analysis discussed here, please see
    sequent violence for the 12–14 age
                                                                                                        Lipsey and Derzon, 1998.
    group—the lack of social ties and
                                                                                                                              continued on next page

   This sidebar is based on “Predictors of Violent or Serious Delinquency in Adolescence and Early Adulthood,” by M.W. Lipsey and J.H. Derzon, in
  Serious and Violent Juvenile Offenders: Risk Factors and Successful Interventions, edited by Rolf Loeber and David P. Farrington (Sage Publications,
  Inc., 1998).

than 17 percent of youth committed a               x Individual:                                               more than doubled the risk that
violent act by age 18, and 80 percent of              y Hyperactivity or attention deficits at                 subjects would engage in violence
them were expected to do so based on                    age 10, 14, or 16 doubled the risk of                  at age 18.
significant predictors seen at age 10.                  violent behavior at age 18.                        y Poor family management prac-
Eighty-four percent were expected to do                                                                      tices and family conflict when
so based on the significant predictors                y Sensation seeking and involvement
                                                        in drug selling at ages 14 and 16                    subjects were age 10 were not
seen at age 16. The results of the Seattle                                                                   significant predictors of later
project are described below for each                    more than tripled the risk of in-
                                                        volvement in violence.                               violence. However, poor family
domain—individual, family, school, peers,                                                                    management practices when
and community and neighborhood                     x Family:                                                 subjects were age 14 doubled
(Herrenkohl et al., in press).                                                                               the risk for later involvement in
                                                      y Parental attitudes favorable to vio-
                                                        lence when subjects were age 10                      violence.

                                                                                             y Low commitment to schooling, low
Table: Ranking of Ages 6–11 and Ages 12–14 Predictors of Violent or                            educational aspirations, and multiple
       Serious Delinquency at Ages 15–25                                                       school transitions at ages 14 and 16
                                                                                               predicted a significantly increased
Predictors at Ages 6–11                            Predictors at Ages 12–14                    risk for involvement in violence at
                                                                                               age 18.
                                      Rank 1 Group
                                                                                           x Peers:
General offenses (.38)                              Social ties (.39)                        y Having delinquent friends at ages
Substance use (.30)                                 Antisocial peers (.37)                     10, 14, and 16 predicted an in-
                                                                                               creased risk for later involvement
                                      Rank 2 Group                                             in violence.
Gender (male) (.26)                                 General offenses (.26)                   y Gang membership at age 14 more
Family socioeconomic status (.24)                                                              than tripled the risk for involvement
Antisocial parents (.23)                                                                       in violence at age 18.
                                                                                             y Gang membership when subjects
                                      Rank 3 Group                                             were age 16 more than quadrupled
                                                                                               the risk for involvement in violence
Aggression (.21)                                    Aggression (.19)
                                                                                               at age 18.
Ethnicity (.20)                                     School attitude/performance (.19)
                                                    Psychological condition (.19)          x Community and neighborhood:
                                                    Parent-child relations (.19)             y Community disorganization, the
                                                    Gender (male) (.19)                        availability of drugs, and knowing
                                                    Physical violence (.18)                    adults involved in criminal activi-
                                                                                               ties at ages 14 and 16 all were asso-
                                      Rank 4 Group                                             ciated with an increased risk for
                                                                                               later involvement in violence.
Psychological condition (.15)                       Antisocial parents (.16)
Parent-child relations (.15)                        Person crimes (.14)
Social ties (.15)                                   Problem behavior (.12)                 Conclusion
Problem behavior (.13)                              IQ (.11)                               More research needs to be done on
School attitude/performance (.13)                                                          youth violence, including studies that
Medical/physical characteristics (.13)                                                     contrast violent offenders and nonvio-
IQ (.12)                                                                                   lent offenders/nonoffenders. Research is
Other family characteristics (.12)                                                         also required to better understand the
                                                                                           protective factors that mitigate the ef-
                                      Rank 5 Group                                         fects of risk exposure. Many predictors
                                                                                           of violent behavior are predictors
Broken home (.09)                                   Broken home (.10)
                                                                                           of other problems, such as substance
Abusive parents (.07)                               Family socioeconomic status (.10)
                                                                                           abuse, delinquency, school dropout, and
Antisocial peers (.04)                              Abusive parents (.09)
                                                                                           teen pregnancy (Dryfoos, 1991; Hawkins,
                                                    Other family characteristics (.08)
                                                                                           Catalano, and Miller, 1992). The risk of
                                                    Substance abuse (.06)
                                                                                           violence is also compounded by the
                                                    Ethnicity (.04)
                                                                                           number of risk factors involved. The
                                                                                           Cambridge Study in Delinquent Develop-
Note: The value in parentheses is the mean correlation between the predictor and the
                                                                                           ment (Farrington, 1997) found that the
outcome, adjusted to equate the source studies on relevant methodological features.
                                                                                           percentage of youth convicted for violent
                                                                                           crimes increased from only 3 percent for
                                                                                           those with no risk factors to 31 percent
                                                                                           for those with four risk factors (low fam-
y Parental criminality when subjects           x School:
                                                                                           ily income, large family size, low nonver-
  were age 14 (not assessed at age 10)             y Low academic performance at           bal IQ at ages 8–19, and poor parental
  more than doubled the risk for in-                 ages 10, 14, and 16 predicted an      childrearing behavior).
  volvement in violence at age 18.                   increased risk for involvement in
y When subjects were age 16, parental                violence at age 18.                   The larger the number of risk factors
  criminality, poor family management,                                                     to which an individual is exposed, the
                                                   y Behavior problems at school (as       greater the probability that the indi-
  family conflict, and residential mobil-            rated by teachers) when subjects
  ity at least doubled the risk for in-                                                    vidual will engage in violent behavior.
                                                     were age 10 significantly predicted   Multicomponent interventions targeting
  volvement in violence at age 18.                   involvement in violence at age 18.    identification of shared predictors and

constellations of risk factors may be          Capaldi, D.M., and Patterson, G.R. 1996.       Chicago, IL: University of Chicago Press,
more effective in preventing violence          Can violent offenders be distinguished         pp. 421–475.
than those that target single risk factors.    from frequent offenders? Prediction from
                                                                                              Farrington, D.P., and Loeber, R. 1999.
                                               childhood to adolescence. Journal of
For more information about this meta-          Research in Crime and Delinquency              Transatlantic replicability of risk factors
analysis, the studies that were examined,                                                     in the development of delinquency. In
and the procedures that were used, see                                                        Historical and Geographical Influences on
Hawkins et al., 1998.                          Catalano, R.F., and Hawkins, J.D. 1996.        Psychopathology, edited by P. Cohen, C.
                                               The social development model: A theory         Slomkowski, and L.N. Robbins. Mahwah,
                                               of antisocial behavior. In Delinquency and     NJ: Lawrence Erlbaum, pp. 299–329.
For Further Information                        Crime: Current Theories, edited by J.D.
                                               Hawkins. New York, NY: Cambridge               Foshee, V., and Bauman, K.E. 1992. Pa-
The following publications are available
                                                                                              rental and peer characteristics as modi-
from the Juvenile Justice Clearinghouse        University Press, pp. 149–197.
                                                                                              fiers of the bond-behavior relationship:
(JJC). For more information or to order a
                                               Denno, D.W. 1990. Biology and Violence:        An elaboration of control theory. Jour-
copy, contact JJC, 800–638–8736 (phone),
                                               From Birth to Adulthood. Cambridge, UK:        nal of Health and Social Behavior
301–519–5600 (fax),
                                               Cambridge University Press.                    33(1):66–76.
(e-mail), (Internet).
                                               Dryfoos, J.G. 1991. Adolescents at risk: A     Hawkins, J.D., Arthur, M.W., and Catalano,
x Summary of Study Group’s Final Re-
                                               summation of work in the field: Programs       R.F. 1995. Preventing substance abuse.
  port. To help communities and practi-
                                               and policies. Journal of Adolescent Health     In Building a Safer Society: Strategic Ap-
  tioners learn more about serious and
                                               12:630–637.                                    proaches to Crime Prevention: Vol. 19,
  violent juvenile offenders, OJJDP re-
                                                                                              Crime and Justice: A Review of Research,
  leased a Bulletin that summarizes the        Elliott, D.S. 1994. Serious violent offend-    edited by M. Tonry and D.P. Farrington.
  Study Group’s final report. The 8-page       ers: Onset, developmental course, and
                                                                                              Chicago, IL: University of Chicago Press,
  Bulletin, Serious and Violent Juvenile       termination—The American Society of            pp. 343–427.
  Offenders (May 1998), is available (free     Criminology 1993 presidential address.
  of charge) from JJC.                         Criminology 32:1–21.                           Hawkins, J.D., Catalano, R.F., and Miller,
x Final Study Group Report. The Study                                                         J.Y. 1992. Risk and protective factors for
                                               Elliott, D.S., Huizinga, D., and Menard, S.
  Group’s final report, Never Too Early,                                                      alcohol and other drug problems in ado-
                                               1989. Multiple Problem Youth: Delinquency,     lescence and early adulthood: Implica-
  Never Too Late: Risk Factors and Suc-        Substance Use and Mental Health Prob-
  cessful Interventions for Serious and                                                       tions for substance abuse prevention.
                                               lems. New York, NY: Springer-Verlag.           Psychological Bulletin 112:64–105.
  Violent Juvenile Offenders (Loeber and
  Farrington, 1997), is also available         Farrington, D.P. 1989. Early predictors of
                                                                                              Hawkins, J.D., Farrington, D.P., and
  (for a fee) from JJC.                        adolescent aggression and adult violence.      Catalano, R.F. 1998. Reducing violence
                                               Violence and Victims 4:79–100.
                                                                                              through the schools. In Violence in Ameri-
References                                     Farrington, D.P. 1991. Childhood aggres-       can Schools: A New Perspective, edited by
                                               sion and adult violence: Early precursors      D.S. Elliott, B.A. Hamburg, and K.R. Will-
Ageton, S.S. 1983. Sexual Assault Among                                                       iams. New York, NY: Cambridge University
Adolescents. Lexington, MA: Lexington          and later-life outcomes. In The Develop-
                                               ment and Treatment of Childhood Aggres-        Press, pp. 188–216.
                                               sion, edited by D.J. Pepler and K.H.           Hawkins, J.D., Herrenkohl, T., Farrington,
Baker, R.L.A., and Mednick, B.R. 1984. In-     Rubin. Hillsdale, NJ: Lawrence Erlbaum,
                                                                                              D.P., Brewer, D., Catalano, R.F., and
fluences on Human Development: A Longi-        pp. 5–29.                                      Harachi, T.W. 1998. A review of predictors
tudinal Perspective. Boston, MA: Kluwer-
                                               Farrington, D.P. 1995. Key issues in the       of youth violence. In Serious and Violent
Nijhoff.                                                                                      Juvenile Offenders: Risk Factors and Suc-
                                               integration of motivational and opportunity-
Battin, S.R., Hill, K.G., Abbott, R.D.,        reducing crime prevention strategies. In       cessful Interventions, edited by R. Loeber
Catalano, R.F., and Hawkins, J.D. 1998.        Integrating Crime Prevention Strategies:       and D.P. Farrington. Thousand Oaks, CA:
The contribution of gang membership to         Propensity and Opportunity, edited by P.O.H.   Sage Publications, Inc., pp. 106–146.
delinquency beyond delinquent friends.         Wikström, R.V. Clarke, and J. McCord.          Henry, B., Avshalom, C., Moffitt, T.E.,
Criminology 36:93–115.                         Stockholm, Sweden: National Council for
                                                                                              and Silva, P.A. 1996. Temperamental and
                                               Crime Prevention, pp. 333–357.                 familial predictors of violent and non-
Brewer, D.D., Hawkins, J.D., Catalano, R.F.,
and Neckerman, H.J. 1995. Preventing se-       Farrington, D.P. 1997. Early prediction of     violent criminal convictions: Age 3 to
rious, violent, and chronic juvenile offend-   violent and nonviolent youthful offending.     age 18. Developmental Psychology
ing: A review of evaluations of selected       European Journal on Criminal Policy and        32:614–623.
strategies in childhood, adolescence, and      Research 5:51–66.                              Herrenkohl, T.I., Maguin, E., Hill, K.G.,
the community. In Sourcebook on Serious,
                                               Farrington, D.P. 1998. Predictors, causes      Hawkins, J.D., and Abbott, R.D. In press.
Violent, and Chronic Juvenile Offenders,                                                      Developmental risk factors for youth
edited by J.C. Howell, B. Krisberg, J.D.       and correlates of male youth violence. In
                                               Youth Violence, Crime and Justice, Vol. 24,    violence. Journal of Adolescent Health.
Hawkins, and J.J. Wilson. Thousand Oaks,
CA: Sage Publications, Inc., pp. 61–141.       edited by M. Tonry and M.H. Moore.

Hirschi, T. 1969. Causes of Delinquency.
Berkeley, CA: University of California        OJJDP Study Group
                                              In 1995, the Office of Juvenile Justice and Delinquency Prevention (OJJDP) convened a
Hogh, E., and Wolf, P. 1983. Violent crime    Study Group on Serious and Violent Juvenile Offenders, a distinguished panel brought
in a birth cohort: Copenhagen 1953–1977.      together to build a research base for policymakers and practitioners who deal with juve-
In Prospective Studies of Crime and Delin-    niles who engage in serious and violent conduct. The group, chaired by Drs. Rolf Loeber
quency, edited by K.T. Van Dusen and S.A.     and David P. Farrington, included 22 leading juvenile justice and criminology scholars
Mednick. Boston, MA: Kluwer-Nijhoff,          selected on the basis of their expert knowledge of different aspects of serious and vio-
pp. 249–267.                                  lent juvenile (SVJ) offending. The OJJDP Study Group documented existing information
                                              about SVJ offenders, examined programs for SVJ offenders, evaluated the programs’
Janosz, M., LeBlanc, M., Boulerice, B., and
                                              performance, and recommended further research and evaluation efforts needed to pre-
Tremblay, R.E. 1996. Disentangling the
                                              vent and control SVJ offending.
weight of school dropout predictors: A
test on two longitudinal samples. Unpub-      The Study Group’s final report, Never Too Early, Never Too Late: Risk Factors and Suc-
lished manuscript.                            cessful Interventions for Serious and Violent Juvenile Offenders, was completed in
                                              1997 under grant number 95–JD–FX–0018. The conclusions of the Study Group were
Kandel, E., and Mednick, S.A. 1991. Perina-
                                              subsequently set forth in a volume entitled Serious and Violent Juvenile Offenders: Risk
tal complications predict violent offend-
                                              Factors and Successful Interventions, edited by the Study Group’s cochairs, Rolf
ing. Criminology 29:519–529.
                                              Loeber and David P. Farrington, and published by Sage Publications, Inc., in 1998.
Klinteberg, B.A., Andersson, T.,
                                              This Bulletin was adapted from two chapters of the published volume. The main text
Magnusson, D., and Stattin, H. 1993. Hy-
                                              was adapted from “A Review of Predictors of Youth Violence,” by J. David Hawkins,
peractive behavior in childhood as re-
                                              Todd I. Herrenkohl, David P. Farrington, Devon Brewer, Richard F. Catalano, and
lated to subsequent alcohol problems and
                                              Tracy W. Harachi. The sidebar text and table were adapted from “Predictors of Vio-
violent offending: A longitudinal study of
                                              lent or Serious Delinquency in Adolescence and Early Adulthood,” by Mark W.
male subjects. Personality and Individual
                                              Lipsey and James H. Derzon.
Differences 15:381–388.
Lipsey, M.W., and Derzon, J.H. 1998. Pre-     Study Group Members
dictors of violent or serious delinquency
in adolescence and early adulthood. In        David M. Altschuler, Ph.D., The Johns Hopkins University, Baltimore, MD.
Serious and Violent Juvenile Offenders:       Alfred Blumstein, Ph.D., Carnegie Mellon University, Pittsburgh, PA.
Risk Factors and Successful Interventions,    Richard F. Catalano, Ph.D., University of Washington, Seattle.
edited by R. Loeber and D.P. Farrington.
Thousand Oaks, CA: Sage Publications,         Julius Debro, Ph.D., University of Washington, Seattle.
Inc., pp. 86–105.                             David P. Farrington, Ph.D., University of Cambridge, England.
Loeber, R. 1990. Development and risk         Peter Greenwood, The RAND Corporation, Santa Monica, CA.
factors of juvenile antisocial behavior and   Nancy G. Guerra, Ph.D., University of Illinois at Chicago.
delinquency. Clinical Psychology Review
                                              Darnell F. Hawkins, Ph.D., University of Illinois at Chicago.
                                              J. David Hawkins, Ph.D., University of Washington, Seattle.
Loeber, R. 1996. Developmental continu-
                                              James C. Howell, Ph.D., Institute for Intergovernmental Research, Tallahassee, FL.
ity, change, and pathways in male juvenile
problem behaviors and delinquency. In         David Huizinga, Ph.D., University of Colorado, Boulder.
Delinquency and Crime: Current Theories,      Barry Krisberg, Ph.D., National Council on Crime and Delinquency, San Francisco, CA.
edited by J.D. Hawkins. Cambridge, UK:
                                              John H. Laub, Ph.D., Northeastern University, Boston, MA.
Cambridge University Press, pp. 1–27.
                                              Marc LeBlanc, Ph.D., University of Montreal, Quebec, Canada.
Loeber, R., and Hay, D.F. 1996. Key issues
in the development of aggression and vio-     Mark W. Lipsey, Ph.D., Vanderbilt University, Nashville, TN.
lence from childhood to early adulthood.      Rolf Loeber, Ph.D., University of Pittsburgh, PA.
Annual Review of Psychology 48:371–410.       Walter B. Miller, Ph.D., Cambridge, MA.
Loney, J., Kramer, J., and Milich, R. 1983.   Mark H. Moore, Ph.D., Harvard University, Cambridge, MA.
The hyperkinetic child grows up: Predic-
                                              Howard N. Snyder, Ph.D., National Center for Juvenile Justice, Pittsburgh, PA.
tors of symptoms, delinquency, and
achievement at follow-up: Birth and child-    Terence P. Thornberry, Ph.D., University at Albany, State University of New York.
hood cohorts. In Handbook of Longitudinal     Patrick H. Tolan, Ph.D., University of Illinois at Chicago.
Research, Vol. 1., edited by S.A. Mednick,
                                              Gail A. Wasserman, Ph.D., Columbia University, New York, NY.
M. Harway, and K.M. Finello. New York, NY:

Maguin, E., Hawkins, J.D., Catalano, R.F.,      at the Society for Prevention Research,          Oaks, CA: Sage Publications, Inc., pp.
Hill, K., Abbott, R., and Herrenkohl, T.        June 1996, San Juan, PR.                         213–237.
1995. Risk factors measured at three ages
                                                Peterson, P.L., Hawkins, J.D., Abbott, R.D.,     Tolan, P.H., and Thomas, P. 1995. The
for violence at age 17–18. Paper presented
at the American Society of Criminology,         and Catalano, R.F. 1994. Disentangling the       implications of age of onset for delin-
                                                effects of parental drinking, family man-        quency risk: II. Longitudinal data. Jour-
November 1995, Boston, MA.
                                                agement, and parental alcohol norms on           nal of Abnormal Child Psychology
Maguin, E., and Loeber, R. 1996. Academic       current drinking by black and white ado-         23:157–181.
performance and delinquency. In Crime           lescents. Journal of Research on Adoles-
and Justice: A Review of Research, Vol. 20,     cence 4:203–227.                                 Wadsworth, M.E.J. 1976. Delinquency,
                                                                                                 pulse rates and early emotional depriva-
edited by M. Tonry. Chicago, IL: Univer-
sity of Chicago Press, pp. 145–264.             Piper, E. 1985. Violent recidivism and           tion. British Journal of Criminology
                                                chronicity in the 1958 Philadelphia co-          16:245–256.
McCord, J. 1979. Some child-rearing ante-       hort. Journal of Quantitative Criminology
cedents of criminal behavior in adult           1:319–344.                                       Wadsworth, M.E.J. 1978. Delinquency pre-
                                                                                                 diction and its uses: The experience of a
men. Journal of Personality and Social
Psychology 37:1477–1486.                        Raine, A., and Jones, R. 1987. Attention,        21-year follow-up study. International Jour-
                                                autonomic arousal, and personality               nal of Mental Health 7:43–62.
McCord, J., and Ensminger, M. 1995. Path-       in behaviorally disordered children.
ways from aggressive childhood to crimi-        Journal of Abnormal Child Psychology             Wells, L.E., and Rankin, J.H. 1988. Direct
                                                                                                 parental controls and delinquency.
nality. Paper presented at the American         15:583–599.
Society of Criminology, November 1995,                                                           Criminology 26:263–285.
                                                Robins, L.N. 1966. Deviant Children Grown
Boston, MA.                                                                                      Widom, C.S. 1989. The cycle of violence.
                                                Up: A Sociological and Psychiatric Study of
McCord, W., McCord, J., and Zola, I.K.          Sociopathic Personality. Baltimore, MD:          Science 244:160–166.
1959. Origins of Crime: A New Evaluation        Williams and Wilkins.                            Wikström, P.O.H. 1985. Everyday Violence
of the Cambridge-Somerville Youth Study.                                                         in Contemporary Sweden. Stockholm:
                                                Rosenthal, R. 1991. Meta-analytic Proce-
New York, NY: Cambridge University                                                               National Council for Crime Prevention.
Press.                                          dures for Social Research (rev. ed.).
                                                Newbury Park, CA: Sage Publications, Inc.        Williams, J.H. 1994. Understanding sub-
Mednick, S.A., and Kandel, E.S. 1988. Con-                                                       stance use, delinquency involvement,
genital determinants of violence. Bulletin      Sampson, R., and Lauritsen, J. 1994. Violent
                                                victimization and offending: Individual-,        and juvenile justice system involvement
of the American Academy of Psychiatry and                                                        among African-American and European-
the Law 16:101–109.                             situational-, and community-level risk
                                                factors. In Understanding and Preventing         American adolescents. Unpublished
                                                                                                 dissertation, University of Washington,
Mitchell, S., and Rosa, P. 1979. Boyhood        Violence: Vol. 3, Social Influences, edited
behavior problems as precursors of crimi-       by A.J. Reiss and J.A. Roth. Washington,         Seattle, WA.
nality: A fifteen-year follow-up study. Jour-   DC: National Academy Press, pp.1–114.            Zingraff, M.T., Leiter, J., Myers, K.A., and
nal of Child Psychology and Psychiatry                                                           Johnson, M. 1993. Child maltreatment and
                                                Smith, C., and Thornberry, T.P. 1995. The
22:19–33.                                                                                        youthful problem behavior. Criminology
                                                relationship between childhood maltreat-
Moffitt, T.E. 1987. Parental mental disor-      ment and adolescent involvement in de-           31:173–202.
der and offspring criminal behavior: An         linquency. Criminology 33:451–481.
adoption study. Psychiatry 50:346–360.                                                           Points of view or opinions expressed in this
                                                Stattin, H., and Magnusson, D. 1989. The         document are those of the authors and do not
Moffitt, T.E. 1993. Adolescent-limited and      role of early aggressive behavior in the         necessarily represent the official position or
life-course-persistent antisocial behavior:     frequency, seriousness, and types of later       policies of OJJDP or the U.S. Department of
A developmental taxonomy. Psychological         crime. Journal of Consulting and Clinical        Justice.
Review 100:674–701.                             Psychology 57:710–718.
Olweus, D. 1979. Stability of aggressive        Thornberry, T.P., Huizinga, D., and Loeber,       The Office of Juvenile Justice and Delin-
reaction patterns in males: A review. Psy-      R. 1995. The prevention of serious delin-         quency Prevention is a component of the Of-
chological Bulletin 86:852–875.                 quency and violence: Implications from the        fice of Justice Programs, which also includes
                                                Program of Research on the Causes and             the Bureau of Justice Assistance, the Bureau
Paschall, M.J. 1996. Exposure to violence
                                                Correlates of Delinquency. In Sourcebook          of Justice Statistics, the National Institute of
and the onset of violent behavior and sub-      on Serious, Violent, and Chronic Juvenile         Justice, and the Office for Victims of Crime.
stance use among black male youth: An
                                                Offenders, edited by J.C. Howell, B. Krisberg,
assessment of independent effects and           J.D. Hawkins, and J.J. Wilson. Thousand
psychosocial mediators. Paper presented

Acknowledgments                          Share With Your Colleagues
J. David Hawkins, Ph.D., is Director     Unless otherwise noted, OJJDP publications are not copyright protected. We
of the Social Development Research       encourage you to reproduce this document, share it with your colleagues, and
Group and Professor of Social Work       reprint it in your newsletter or journal. However, if you reprint, please cite OJJDP
at the University of Washington,         and the authors of this Bulletin. We are also interested in your feedback, such as
Seattle, and a founder of Develop-       how you received a copy, how you intend to use the information, and how OJJDP
mental Research and Programs.            materials meet your individual or agency needs. Please direct your comments and
                                         questions to:
Todd I. Herrenkohl, Ph.D., is a
Research Analyst with the Social                                 Juvenile Justice Clearinghouse
Development Research Group and                                   Publication Reprint/Feedback
Assistant Professor of Social Work at                            P.O. Box 6000
the University of Washington, Seattle.                           Rockville, MD 20849–6000
David P. Farrington, Ph.D., is Profes-                           301–519–5600 (fax)
sor of Psychological Criminology at                              E-Mail:
the Institute of Criminology, Univer-
sity of Cambridge, England.
Devon Brewer, Ph.D., is a Research
Scientist with the Alcohol and Drug
Abuse Institute at the University of
Washington, Seattle.
Richard F. Catalano, Ph.D., is
Associate Director of the Social
Development Research Group and
Professor of Social Work at the
University of Washington, Seattle,
and a founder of Developmental
Research and Programs.
Tracy W. Harachi, Ph.D., is a Re-
search Associate Professor with the
Social Development Research Group
in the School of Social Work at the
University of Washington, Seattle.
Lynn Cothern, Ph.D., is a Senior
Writer/Editor for the Juvenile Justice
Resource Center in Rockville, MD.
All photographs copyright © 1999
PhotoDisc, Inc.

U.S. Department of Justice                              PRESORTED STANDARD
Office of Justice Programs                               POSTAGE & FEES PAID
Office of Juvenile Justice and Delinquency Prevention      PERMIT NO. G–91

Washington, DC 20531
Official Business
Penalty for Private Use $300

                     Bulletin                              NCJ 179065

To top