psychosocial characteristics of young violent offender

10
Psychosocial characteristics of young violent offenders: a comparative study ROGER C. KATZ AND JOE MARQUETTE University of the Pacific, Stockton, California, USA ABSTRACT When young people commit violent crimes such as homicide, the act is sometimes attributed to poorly controlled anger, rage or paranoid symptomatology. To test this assertion, we hypothesised that young men incarcerated for murder would show more anger, hostility and paranoid ideation, as well as increased levels of global psychopathology, than a cohort of non-violent offenders and normal high school students. These predictions were not supported. Young murderers were indis- tinguishable from the two control groups on the MMPI-A, the State-Trait Anger Scale, and the SCL-90-R. Moreover, they showed no evidence of increased anger, paranoid features or global psychopathology relative to standardisation samples. Rather than being differentiated by personality characteristics, the data suggest that a history of gang membership and violent behaviour do a better job of discriminating between these groups. INTRODUCTION Although the incidence of murder and other violent crime is declining in the United States, violent crime among young people is increasing at an alarming rate (Blumstein, 1995; Snyder & Sickmund, 1995). This is true of young per- petrators as well as young victims. Unless there is a significant improvement in crime prevention, the upsurge in violence among young Americans is expected to continue well into the next decade reflecting a projected increase in the crime-prone 16- to 24-year-old age-group. According to the US Department of Justice, juveniles are responsible for one in five violent crimes in the USA (Snyder & Sickmund, 1995). They commit an excessive number of murders as evidenced by the fact that the murder rate among young people more than doubled between 1984 and 1991. In contrast, the murder rate among adults increased only 20% during the same period. Most juvenile murderers (91%) are males. A disproportionate number of them are black. Handguns are their weapon of choice, and 30% of their Criminal Behaviour and Mental Health, 6, 339–348 1996 © Whurr Publishers Ltd 339

Upload: sarah-white

Post on 11-Apr-2016

8 views

Category:

Documents


2 download

DESCRIPTION

Características psicossociais e jovens violentos

TRANSCRIPT

Psychosocial characteristics ofyoung violent offenders: acomparative study

ROGER C. KATZ AND JOE MARQUETTE University of the Pacific,Stockton, California, USA

ABSTRACT When young people commit violent crimes such as homicide, the actis sometimes attributed to poorly controlled anger, rage or paranoid symptomatology.To test this assertion, we hypothesised that young men incarcerated for murderwould show more anger, hostility and paranoid ideation, as well as increased levels ofglobal psychopathology, than a cohort of non-violent offenders and normal highschool students. These predictions were not supported. Young murderers were indis-tinguishable from the two control groups on the MMPI-A, the State-Trait AngerScale, and the SCL-90-R. Moreover, they showed no evidence of increased anger,paranoid features or global psychopathology relative to standardisation samples.Rather than being differentiated by personality characteristics, the data suggest that ahistory of gang membership and violent behaviour do a better job of discriminatingbetween these groups.

INTRODUCTION

Although the incidence of murder and other violent crime is declining in theUnited States, violent crime among young people is increasing at an alarmingrate (Blumstein, 1995; Snyder & Sickmund, 1995). This is true of young per-petrators as well as young victims. Unless there is a significant improvementin crime prevention, the upsurge in violence among young Americans isexpected to continue well into the next decade reflecting a projected increasein the crime-prone 16- to 24-year-old age-group.

According to the US Department of Justice, juveniles are responsible forone in five violent crimes in the USA (Snyder & Sickmund, 1995). Theycommit an excessive number of murders as evidenced by the fact that themurder rate among young people more than doubled between 1984 and 1991.In contrast, the murder rate among adults increased only 20% during the sameperiod. Most juvenile murderers (91%) are males. A disproportionate numberof them are black. Handguns are their weapon of choice, and 30% of their

Criminal Behaviour and Mental Health, 6, 339–348 1996 © Whurr Publishers Ltd 339

CBMH 6(4) CRC 15/12/05 11:06 am Page 339

victims are strangers who are killed during the commission of a felony(Cornell, 1993; Snyder & Sickmund, 1995).

Whilst there is an abundance of theory to explain why young people com-mit violent crimes, including the most violent crime, murder, there is noagreed cause or group of causes that explains all instances of this behaviour.The consensus seems to be that criminal violence occurs for different reasonsin different people. Some theories point to a genetic predisposition to crimi-nality (Brennan & Mednick, 1993), or to neurological abnormalities such asgeneralised or partial complex seizures (Lewis, Shanok, Grant & Ritvo, 1983;Lewis, Moy, Jackson, Aaronson, Restifo, Serra & Simos, 1985; Pincus &Tucker, 1985). According to Lewis and Pincus (1989), repeated acts of vio-lence in combination with central nervous system injury, limbic and otherneurological dysfunction, cognitive impairment, episodic psychosis, and a his-tory of abuse and family violence is a profile that fits many violent young men.Schizophrenia and excessive paranoid ideation have been observed in someadolescent murderers (Lewis, 1976; Sendl & Blomgren, 1975; Mouridsen &Tolstrup, 1988). For psychodynamic theorists, homicidal behaviour has beenattributed to unconscious conflicts, an ego weakness, displaced anger, or toregressive coping mechanisms for dealing with extreme stress (Sheley, 1985;Atlas & Porzio, 1994)). Unfortunately, such theories are hard to test becauseof their ambiguity. More easily measured psychosocial variables are frequentlycited as contributors to teenage violence (Busch, Zagar, Hughes, Arbit &Russell, 1990; Labelle, Bradford, Bourget, Jones & Carmichael, 1991; Zagar,Arbit, Sylvies, Busch & Hughes, 1990). Among them are poverty, family tur-moil, gang involvement, excessive alcohol intake, physical abuse, educationaldifficulties and exposure to violent role models, either within the family orthrough the media (Bandura, 1973). In line with this view, the USDepartment of Justice notes that the current epidemic of teenage violencecoincides with the proliferation of drug trafficking, guns and violent streetgangs which began in the mid-1980s (Snyder & Sickmund, 1995).

Empirically based studies of psychological factors that predispose youngmen to violence are few in number. This is especially true of studies with rela-tively large sample sizes and appropriate control groups that use psychometri-cally sound instruments to assess specific personality attributes.

The present study investigated anger proneness, hostility and paranoidideation as correlates of violent behaviour in young men. Although excessiveanger and paranoia have been linked to violent outbursts (Lewis, 1976; Atlas& Porzio, 1994), this is the first time these attributes have been examined in acontrolled study. The sample was limited to males because most young violentoffenders are males. Like Goldstein and Glick (1987) and others (Dodge,1985; Lewis, Lovely, Yeager, Ferguson, Friedman, Sloane, Friedman & Pincus,1988), it was felt that excessive anger and hostility, distorted attributions ofthreat, interpersonal alienation, heightened sensitivity to provocation anddeficits in anger control skills are factors that predispose violent behaviour. It

340 Katz, Marquette

CBMH 6(4) CRC 15/12/05 11:06 am Page 340

was also believed that emotional distress and maladjustment are potentiatingvariables for violent behaviour. To test this hypothesis, a battery of psychome-tric tests was administered to three groups of young men. One group consistedof adjudicated delinquents who were incarcerated in the California YouthAuthority for first- or second-degree murder. Participants in the second groupwere also wards of the California Youth Authority but placed there for a non-violent crime. The third group was a control group of high school juniors andseniors with no significant criminal history. It was predicted that young mur-derers would show more anger, hostility, paranoid thinking and global psy-chopathology than the other two groups.

METHOD

Subjects

Subjects (n = 77) were males ranging in age from 16 to 23 (mean = 19.01, SD = 2.09). All of them participated voluntarily. They were told the purpose ofthe study was to assess various attitudes and beliefs in young people and thatthey would receive US$5 for their participation. Answers were given anony-mously. Data were collected in a group setting. In all cases, the subjects wereencouraged to be as candid as possible in completing the questionnaires. Twoof the groups were recruited from the California Youth Authority in Stockton,CA. One group (n = 29) consisted of convicted first- and second-degree mur-derers. The other group was made up of non-violent offenders (n = 15) whowere incarcerated because of repeated acquisitive crimes and/or violation ofprobation. Approval to conduct the study was obtained from the HumanResearch Committee of the California Youth Authority in Sacramento.Subjects in both Youth Authority groups received $5 for their participation.The third group (n = 33) consisted of high school juniors and seniors from apublic high school in Stockton, CA.1 These students were enrolled in a psy-chology class where they received extra credit for their participation. Subjectsin all three groups were ethnically diverse and consisted of roughly equal num-bers of Caucasians, Blacks, Hispanics and Asians. The two Youth Authoritygroups had a somewhat larger number of Blacks (27%) and Hispanics (38%).

Measures

Data were collected using a brief demographic questionnaire, along with threestandardised psychometric measures (see below), each of which has acceptablereliability and validity. The demographic questionnaire, developed by the

Psychosocial characteristics of young violent offenders 341

1 The mean age for the three groups was 20.6 for the murderers, 19.4 for the non-violent offend-ers, and 16.6 for the high school students. Whilst both Youth Authority groups were older thanthe high school sample, their average age at the time of their committing offence would have beenseveral years younger.

CBMH 6(4) CRC 15/12/05 11:06 am Page 341

authors, used Likert scales (with anchors of 0 and >10) to assess the number oftimes the participants were arrested and the number of times they were arrestedfor a violent crime. Respondents were also asked if they were, or had been, agang member, and if anyone else in their family had been incarcerated for acrime.

The data were collected during a single session that lasted between twoand three hours. Each group was tested individually. If a respondent had diffi-culty reading, help was provided. Respondents were asked not to identifythemselves by name, to encourage honest reporting. The measures aredescribed below.

MMPI-A

The MMPI-A (Butcher, Williams, Graham, Archer, Tellegen, Yossef &Kaemmer, 1992) is a revision of the original MMPI for use in the psychologi-cal evaluation of adolescents. Although the MMPI has been used before in astudy of this type (Cornell, Miller & Benedek, 1988; Truscott, 1993), theMMPI-A has not. The MMPI-A consists of 478 true–false items which yieldscores on 10 clinical scales, as well as several content and supplemental scalesunique for adolescents. For purposes of the study, all of the clinical scales wereanalysed along with four content scales: adolescent anger, family problems,alienation and conduct problems. Like the MMPI-2, T-scores >65 are consid-ered clinically significant. T-scores in the 60–64 range are considered moder-ate elevations which can be interpreted but with less confidence.

SCL-90-R

The SCL-90-R (Derogatis, 1994) is a 90-item self-report measure of currentsymptom status and global distress. Each item is rated on a five-point scale ofdistress (0–4) ranging from ‘not at all’ to ‘extremely’. Raw scores are convertedto T-scores. T-scores >63 are considered clinically meaningful. The SCL-90-Ris usually scored and interpreted in terms of nine primary symptom dimensionsand three global indices of distress. For purposes of this study, however, not allof the symptom dimensions were analysed. Our focus was on three of thedimensions – hostility, paranoid ideation and interpersonal sensitivity – alongwith the global severity index (GSI) which provides a measure of the ‘average’level of distress. The GSI is the most sensitive single indicator of the respon-dent’s psychological status.

State-Trait Anger Scale2 (STAS)

The STAS (Speilberger, Jacobs, Russel & Crane, 1983) is a self-report mea-sure that assesses anger as an emotional state which varies in intensity, as well

342 Katz, Marquette

2 This instrument exists in a 30-item form and a shortened 20-item form. The short form wasused in this study. Both forms have good internal reliability and validity.

CBMH 6(4) CRC 15/12/05 11:06 am Page 342

as a relatively stable personality trait. State anger is defined as an emotionalcondition consisting of feelings of tension, annoyance, irritation or rage. Traitanger, on the other hand, is defined in terms of how frequently a person feelsstate anger over time. A person high in trait anger is likely to perceive moresituations as anger provoking and respond with high trait-anger scores. Eachitem on the scale is rated on a four-point Likert scale from ‘not at all’ (1) to‘very much so’ (4). Higher scores are associated with greater anger.

RESULTS

Mean raw and T-scores for the dependent measures are given in Table 1.Univariate analyses of variance were performed on the group means. This wasfollowed by t-tests to examine planned comparisons and post hoc Scheffe orchi-square tests to examine any other differences between the groups.

Psychosocial characteristics of young violent offenders 343

TABLE 1: Mean raw scores for each dependent measure

Murderers Non-violent offenders High school students

MMPI-A:L 3.53 (52) 3.6 (53) 2.96 (49)K 11.82 (50) 14.06 (53) 12.32 (51)F 9.29 (49) 10.06 (50) 10.0 (50)Hs 7.73 (50) 7.07 (49) 9.56 (55)D 19.58 (50) 22.8 (54) 21.2 (52)Hy 18.94 (48) 23.54 (56) 21.72 (52)Pd 23.77 (56) 25.73 (62) 22.68 (53)Pa 14.53 (52) 14.73 (52) 14.96 (53)Pt 17.74 (48) 18.47 (50) 20.88 (54)Sc 22.17 (49) 26.93 (52) 24.44 (51)Ma 22.71 (51) 23.07 (52) 24.84 (54)Si 25.23 (49) 24.66 (48) 22.32 (45)Anger 7.94 (53) 8.73 (54) 7.60 (53)Conduct 12.71 (56) 12.47 (55) 11.36 (52)Alienation 6.15 (49) 6.13 (49) 6.44 (50)Family problems 11.56 (49) 13.47 (52) 12.72 (50)STAS:state 11.91 11.66 12.32 trait 17.15 20.13 20.08 SCL-90-R:Hostility 0.68 (49) 0.83 (51) 0.67 (49) Interpersonal 0.83 (46) 0.69 (44) 0.77 (45)sensitivityParanoia 1.07 (52) 1.04 (52) 1.03 (52)Global Severity 0.74 (45) 0.77 (45) 0.85 (46)Index

Note: Corresponding T-scores are shown in parentheses

CBMH 6(4) CRC 15/12/05 11:06 am Page 343

The prediction that young murderers would show increased levels of anger,hostility, paranoid ideation, feelings of alienation and global psychopathologywas not supported. There were no significant between-group differences onany of the psychometric measures. As evidenced by their T-scores, the scoresof our sample were very similar to the normative groups on which the testswere standardised. The mean scores for all three groups were ‘within normallimits’.3

Analysis of the questionnaire data showed that more murderers were gangmembers (65% vs 26% and 0% for non-violent offenders and high school stu-dents, respectively), chi square (df = 2) = 27.06, P <0.0001, who had morearrests for violent crime, chi square (df = 8) = 61.55 , P <0.0001. Non-violentoffenders, however, had the most arrests overall, chi square (df = 10) = 75.39,P <0.0001. Although 36% of the high school sample had previously beenarrested, none of them was ever arrested for a violent crime. Compared withhigh school students, a significantly higher proportion of the Youth Authoritysample grew up in a family where at least one other family member spent timein jail or prison, chi square (df = 2) = 10.20, P = 0.006. This was true of 74%of the murderers, 60% of the non-violent offenders, and 32% of the highschoolers.

A significant interaction was observed on the Global Severity Index of theSLC-90-R between group belonging and having a family member convicted ofa crime, F (2, 67) = 3.20, P = 0.04. The interaction was true of non-violentoffenders and high school students but not murderers. Among high school stu-dents, a family history of criminal behaviour was associated with increasedlevels of psychological disturbance. In the non-violent offender group, theeffect of having a family member convicted of a crime was just the opposite –i.e. less distress and fewer symptoms. There was no relationship between therespondent’s age and distress level (rxy = –0.08). Thus, age differences cannotaccount for these findings.

A discriminant function analysis was performed to select the combinationof variables that best predicted membership in the three groups. Ten variableswith the highest Wilkes lambda values were included in the analysis. Twofunctions were calculated with a combined chi square (df = 20) = 146.99,P <0.0001. Removal of the first function still resulted in highly significant dis-criminating power, chi square (df = 9) = 67.73, P <0.001. The two functionsaccounted for 57% and 43%, respectively, of the between-group variability.The first function maximally separated high school students from murderersand non-violent offenders, whilst the other function separated murderers fromnon-violent offenders. The percentage of cases correctly classified by the func-tions was 96% which is approximately 200% better than chance (i.e. 33%).

344 Katz, Marquette

3 On the Pd scale of the MMPI-A, the mean score of the non-violent offender group was in themoderately elevated range (T = 62). This was the most extreme deviation from ‘normal’ thatwe observed.

CBMH 6(4) CRC 15/12/05 11:06 am Page 344

Murderers, compared with non-violent offenders, were more likely to havearrests for violent crime but to have fewer arrests overall. They were morelikely to be gang members and more inclined to express feelings of self-con-sciousness, inferiority, and inadequacy. Compared with high school students,the young offenders expressed less global distress and seemed less troubled byfeelings of anger and hostility. It was also less common for the students to bearrested for a crime, to be a gang member, or to come from a family whereantisocial behaviour was a problem.

DISCUSSION

Our results do not support the contention of other investigators (Lewis, 1976;Atlas and Porzio, 1994) that violent acts committed by young men are an out-growth of poorly modulated anger and/or paranoid-like symptoms. As a group,the murderers in this study were free of psychopathology based on the mea-sures we used. They did not differ from non-violent offenders on these mea-sures, nor from a cohort of high school students with no significant criminalhistory. There was no evidence that they were any more angry, hostile, psy-chotic, paranoid, psychopathic or emotionally unstable and distressed thanthe young men in the other two groups. Like them, the mean scores of theyoung murderers were well within normal limits in comparison with standard-isation samples.

Other studies have also found no differences between murderers andnon-violent offenders on the MMPI (Cornell, Miller & Benedek, 1988;McDonald & Paitich, 1981; Truscott, 1993) as well as other indices ofpsychopathology (Yates, Beutler & Crago, 1983; Toupin & Morissette,1990). This observation can now be extended to include the MMPI-A aswell as state and trait measures of anger and a point-in-time measure ofsymptom distress, the SCL-90-R. In the same vein, the results of previousstudies (Cornell, Benedek & Benedek, 1987; Busch et al., 1990; Zagar et al.,1990; Truscott, 1993) suggest that differences between young murderers andnon-violent offenders have less to do with mental health problems and moreto do with sociocultural, behavioural and environmental variables. This wastrue in the present study as well. Where the two groups differed was along adimension of gang membership and having a history of violent behaviour.The murderers in this study fell into this category much more often thanyoung men in the other two groups.

An interaction was observed between group membership and having afamily member who was incarcerated for a crime. When this occurred amonghigh schoolers, it was associated with increased levels of psychological distur-bance, especially feelings of anxiety, depression and alienation. When it wasnot the case, distress levels in this group were relatively low. Among non-violent offenders, a family history of criminality had just the opposite effect,i.e. less distress and fewer symptoms in comparison with cohorts without such

Psychosocial characteristics of young violent offenders 345

CBMH 6(4) CRC 15/12/05 11:06 am Page 345

a history. The scores of young murderers were unaffected by this variable. Oneimplication of this finding is that non-violent offenders from criminally orien-tated families may view their behaviour as ‘just the same’ as that of other fam-ily members and as a result they may be less distressed by what they did. It alsoraises the possibility that their criminal behaviour was patterned after antiso-cial role models within the family. For young offenders from less deviant fami-lies, criminal behaviour may reflect the workings of another set of variables;for example, problems brought on by the use of drugs, alcohol, school failure,or alienation and rebelliousness (Busch et al., 1990; Hawkins, 1995).

The generalisability of these findings is open to question. First, many ofthe murderers in the study were incarcerated for years and received coun-selling during this time. Maturational processes and the counselling theyreceived could have mitigated psychological problems that were presentwhen their crime was committed or when they first arrived at the institu-tion. Our findings are only generalisable to young murderers who were adju-dicated through the juvenile justice system and subsequently housed in ajuvenile correction facility. Another point concerns the heterogeneity of themurderers we assessed. What these young men have in common is that theycommitted homicide, which is not to say they killed someone for the samereasons or under the same circumstances. In a study similar to ours, Cornell,Miller and Benedek (1988) found no differences on the MMPI between agroup of young murderers and a control group of delinquents who wereincarcerated for larceny. However, when the MMPI profiles of two sub-groups of murderers were compared – those who committed homicide duringanother crime (e.g. robbery) and those who killed during an interpersonalconflict – the young men in the former group had significantly higher scoreson several of the scales (F, Hs, Hy, Sc). We were unable to control for thisvariable. If we had done so, the results might have been different.

In summary, although young men commit violent crime for many reasons,we found no evidence that anger, hostility and other indices of global psy-chopathology, as measured by three commonly used psychometric instru-ments, differentiated young murderers from non-violent offenders or from nor-mal high school students. A history of gang membership and violent behav-iour did a better job of distinguishing between these groups. We found no evi-dence of significant psychopathology in any of the groups we surveyed. This isin keeping with a recent report from the US Department of Justice(Blumstein, 1995; Snyder & Sickmund, 1995) which suggests that sociocul-tural variables are the driving force behind the country’s rising tide of juvenilecrime and violence.

ACKNOWLEDGEMENTS

Appreciation is extended to Drs Coby Ward and Roseann Hannon, EricMayville and officials at the California Youth Authority for their support in

346 Katz, Marquette

CBMH 6(4) CRC 15/12/05 11:06 am Page 346

this project. The study was partly funded by a grant from the University of thePacific, Mr James Wilson and Syntex Pharmaceuticals.

REFERENCES

ATLAS, J. & PORZIO, L. (1994). Rage and anger: dealing with the symptoms of dysfunction in cur-rent American society. Journal of Psychhistory 22, 103–113.

BANDURA, A. (1973). Aggression: A Social Learning Analysis. Englewood Cliffs, NJ: Prentice Hall.BLUMSTEIN, A. (1995). Violence by young people: why the deadly nexus? National Institute of

Justice Journal (August), 2–9.BRENNAN, P. & MEDNICK, S. (1993). Genetic perspectives on crime. Acta Psychiatrica Scandinavica

87(Suppl. 370), 19–26.BUSCH, K., ZAGAR, R., HUGHES, J., ARBIT, J. & RUSSELL, R. (1990). Adolescents who kill. Journal of

Clinical Psychology46, 472–485.BUTCHER, J., WILLIAMS, C., GRAHAM, J., ARCHER, R., TELLEGEN, A., YOSSEF, S.& KAEMMER, B. (1992).

Minnesota Multiphasic Personality Inventory – Adolescent. Minneapolis, MN: University ofMinnesota Press.

CORNELL, D. (1993). Juvenile homicide: a growing national problem. Behavioral Sciences & theLaw 11, 389–396.

CORNELL, D., BENEDEK, E. & BENEDEK, D. (1987). Characteristics of adolescents charged withhomicide: review of 72 cases. Behavioral Sciences & the Law 5, Special Issue: HomicidalBehavior, 11–23.

CORNELL, D., MILLER, C. & BENEDEK, E. (1988). MMPI profiles of adolescents charged with homi-cide. Behavioral Sciences & the Law 6 Special Issue: Juvenile Delinquency, 401–407.

DEROGATIS, L. (1994). Symptom Checklist-90-R. Minneapolis, MN: National Computer Systems.DODGE, K. (1985). Attributional bias in aggressive children. In: P. Kendall (Ed.), Advances in

Cognitive Behavioral Research and Therapy, Vol. 4. Orlando, FL: Academic Press.GOLDSTEIN, A. & GLICK, B. (1987). Aggression Replacement Training: A Comprehensive Intervention

for Aggressive Youth. Champaign, IL: Research Press.HAWKINS, J. (1995). Controlling crime before it happens: risk-focused prevention. National

Institute of Justice Journal (August), 10–18.LABELLE, A., BRADFORD, J., BOURGET, D., JONES, B. & CARMICHAEL, M. (1991). Adolescent murder-

ers. Canadian Journal of Psychology 36, 583–587.LEWIS, D. (1976). Delinquency, psychomotor epileptic symptomatology, and paranoid sympto-

matology: a triad. American Journal of Psychiatry 133, 1395–1398.LEWIS, D., LOVELY, R., YEAGER, C., FERGUSON, G., FRIEDMAN, M., SLOANE, G., FRIEDMAN, H. & PINCUS,

J. (1988). Intrinsic and environmental characteristics of juvenile murderers. Journal of theAmerican Academy of Child and Adolescent Psychiatry 27, 582–587.

LEWIS, D. MOY, E., JACKSON, D., AARONSON, R., RESTIFO, N., SERRA, S. & SIMOS, A. (1985).Biopsychosocial characteristics of children who later murder: a prospective study. AmericanJournal of Psychiatry 142, 1161–1166.

LEWIS, D. & PINCUS, J. (1989). Epilepsy and violence: evidence for a neuropsychiatric-aggressivesyndrome. Journal of Neuropsychiatry & Clinical Neurosciences 1, 413–418.

LEWIS, D., SHANOK, S., GRANT, M. & RITVO, E. (1983). Homicidally aggressive young children:neuropsychiatric and experiential correlates. American Journal of Psychiatry 140, 148–153.

MCDONALD, A. & PAITICH, D. (1981). A study of homicide: the validity of predictive test factors.Canadian Journal of Psychiatry 26, 549–554.

MOURIDSEN, S. & TOLSTRUP, K. (1988). Children who kill: a case study of matricide. Journal ofChild Psychology and Psychiatry 29, 511–515.

PINCUS, J. & TUCKER, G. (1985). Behavioral Neurology. New York: Oxford University Press.

Psychosocial characteristics of young violent offenders 347

CBMH 6(4) CRC 15/12/05 11:06 am Page 347

SENDLE, I. & BLOMGREN, O. (1975) A comparative study of predictive criteria in the predisposi-tion of homicidal adolescents. American Journal of Psychiatry 132, 423–427.

SHELEY, J. (1985). America’s Crime Problem: An Introduction to Criminology. Belmont, CA:Wadsworth.

SNYDER, H. & SICKMUND, M. (1995). Juvenile Offenders and Victims: A National Report.Washington, DC: Office of Juvenile Justice and Delinquency Prevention.

SPEILBERGER, C., JACOBS, G., RUSSEL, S. & CRANE, R. (1983). Assessment of anger: the state-traitanger scale. In: J. Butcher & C. Spielberger (Eds), Advances in Personality Assessment, Vol.2, pp.159–187. Hillsdale, NJ: Erlbaum.

TOUPIN, J. & MORISSETTE, L. (1990). Juvenile homicide: a case control study. Medicine and theLaw 9, 986–994.

TRUSCOTT, D. (1993). Adolescent offenders: comparison of sexual, violent, and propertyoffences. Psychological Reports 73, 657–658.

YATES, A., BEUTLER, L. & CRAGO, M. (1983). Characteristics of young violent offenders. Journal ofPsychiatry and the Law (Summer), 137–149.

ZAGAR, R., ARBIT, J., SYLVIES, R., BUSCH, K. & HUGHES, J. (1990). Homicidal adolescents: a replica-tion. Psychological Reports 67, 1235–1242.

Address correspondence to: Roger C. Katz, Psychology Department, University of the Pacific,Stockton, CA 95211, USA.

348 Katz, Marquette

CBMH 6(4) CRC 15/12/05 11:06 am Page 348