|
Abstract
|
30 homicidal delinquents were matched with 30 nonviolent delinquents on age, race, sex, and socioeconomic status. Both groups received physical, psychological, educational, psychiatric and social examinations. Data were evaluated using stepwise discriminant analysis. Previous research was replicated. Homicidal adolescents shared four symptoms: criminally violent families, gang participation, alcohol abuse, severe learning difficulties. In addition, when the sample and the original group of murderers were combined the contrast between groups became clearer. Juvenile murderers lived in criminally violent families learning aggression during early childhood. Cognitive and perceptual deficits associated with epilepsy and central nervous system (CNS) conditions led to difficulties in the school years. During adolescence, gang participation and alcohol abuse added to earlier risk factors.
VioLit summary
OBJECTIVE:
The objective of this study by Zagar et al. was to replicate an earlier study in 1990 (Busch, Zagar, Arbit, Hughes and Bussell) comparing juvenile murderers with matched nonviolent delinquents. The null hypothesis: Adolescents who kill do not differ significantly from matched nonviolent delinquents in the presence of criminally violent families, severe educational problems, alcohol abuse and participation in gangs, when using the discriminant analysis coefficients of the predictor variables from the study done by Busch et al. in 1990.
METHODOLOGY:
A quasi-experimental group matched design was used, comparing 30 juveniles convicted of homicide with 30 nonviolent juvenile delinquents. They were matched by age, race, sex and socioeconomic status. The homicidal group represented all delinquents charged with homicide from the sample of 2,016 adjudicated delinquents. This sample of 2,016 was from a larger population of 42,655 adolescents, which is about 5% of the adjudicated delinquent population. These 2,016 all had complete records and had been referred for physical, psychological, educational, psychiatric and social examinations. Thus the sample in this post hoc study was retrospective and not random, reflecting selection bias. There was also experimenter bias reflected in the examiners' knowledge of the murderers. The dependent variables included the following: 1) Physical examinations, with results coded by the International Classification of Diseases (ICD-9), with special attention to Central Nervous System (CNS) disorders, psychological disturbances, substance abuse, and neonatal conditions. Hospital records and neonatal records were reviewed. 2) Psychological and education examinations consisting of several tests: Weschler Intelligence Scale for Children, Bender Visual-motor Gestalt test scored by the Koppitz method for errors with 2 independent raters, Gates-MacGinitie Reading Tests, and the Stanford Achievement Tests. Behavior disorders were classified by the major categories of mental retardation, attention deficit-hyperactivity disorder, attention deficit disorder undifferentiated, and developmental delays (reading, language and arithmetic). There was high correlation between observers for the retardation, ADHD and ADD categories (.94(p<.01), .92(p<.01) and .93(p<.01). 3). Psychiatric examinations of child, adolescent, parents and relatives or guardians consisting of histories totalling approximates 1-4 hours. 4) Social examinations by probation officers supplemented by social workers compiled with descriptions of offenses, family composition, school history and demographics. The nominal data were coded in the areas of violent family members (from court records), physical abuse (documented by physicians, State Dept of Children and Family Services, investigators and police officers), gang participation, substance abuse, and diagnostic category from psychological and educational examinations. Pearson r coefficients were acceptable between observers (between .75 and .87 (p<.01) for the ratings. Stepwise discriminant analysis was used to compare the groups, since many of dependent variables tended to be highly correlated. The discriminant function coefficients or weights that were developed from the earlier study were used in this study. The groups from this study were then combined with the groups from the earlier study (71 in each group from the earlier study), and discriminant analysis was then done on these combined groups and compared with the results of the earlier study by Busch et al. in 1990.
FINDINGS/DISCUSSION:
The variables which contributed to the significant differences (p<.05) between groups were 1) criminally violent families, 2) participation in gangs, 3) alcohol abuse and 4) severe educational difficulties. This replicated the results of the earlier study, and when the groups from the two studies were combined, these differences were even more significant with the larger N's. It was summarized that these juveniles lived in criminally violent families and learned aggressive responses early in childhood. Their cognitive and perceptual difficulties led to problems in school, and alcohol abuse and gang participation in adolescence added to the early risk factors.
AUTHORS' RECOMMENDATIONS:
Larger samples of adolescents would indicate more precisely the differential contributions of the variables in predicting homicidal adolescents.
EVALUATION:
Since this was a post hoc study with very narrow selection criteria--that of participants having complete records and having been tested in all the desired areas including physical, psychological, educational, psychiatric and social examinations--the results are not generalizable to the total population of delinquents. It might be interesting to try to duplicate these studies with a study of the same dependent variables, using random selection from this population of delinquents, and then administering the tests. (CSPV Abstract - Copyright © 1992-2007 by the Center for the Study and Prevention of Violence, Institute of Behavioral Science, Regents of the University of Colorado)
N1 - Call Number: F-33, AB-33
KW - Juvenile Offender
KW - Juvenile Violence
KW - Juvenile Homicide
KW - Homicide Offender
KW - Psychosocial Factors
KW - Research Replication
KW - Family Criminality
KW - Family History
KW - Family Characteristics
KW - Homicide Causes
KW - Violence Causes
KW - Violent-Nonviolent Comparison
KW - Juvenile Delinquency
KW - Juvenile Gang
KW - Gang Effects
KW - Demographic Factors
KW - Juvenile Substance Use
KW - Offender Substance Use
KW - Substance Use Effects
KW - Alcohol Use Effects
KW - Learning Disability
KW - Offender Characteristics
KW - Sociocultural Factors
KW - Child Development
KW - Developmental Pathway
KW - Juvenile Development
Language: en |