Psychosocial variables associated with delinquency in incarcerated adolescents: a comparison by gender and ethnicity

Abstract

This study described and compared psychosocial variables associated with delinquency in a sample of incarcerated adolescents. Comparisons were made across gender and ethnicity. Participants were Mexican-American, Spanish-American, and White non-Hispanic juvenile offenders who had recently been incarcerated. Data regarding delinquency and psychosocial characteristics were collected through written surveys. Psychosocial variables examined were related to family background, peer characteristics, drug use, school liking and success, psychological adjustment, deviance, and histories of sexual or physical trauma/abuse. There was also an examination o f agreement between self-reported drug use and the results of toxicology screens (urine analysis). Data analysis included an exploration of the variables and their relationships to gender and ethnic identification. Study hypotheses included two predictions developed from a recently proposed model of a developmental pathway to delinquency for girls, as well as several hypotheses based on a previous related investigation.

Responses to representative survey items were first examined in order to gain a preliminary understanding of the aggregated personal histories of participants. Mean responses indicated participants generally felt close to their families, and lived with parents. Parental divorce was reported by nearly half of respondents. The average respondent reported fairly strong sanctions against drug use in the family. A majority reported severe school problems, including having flunked a year, and reported many markers of delinquency among their best friends, such as having stolen a car and been incarcerated. The average respondent showed a lukewarm to cool regard for religion, and a surprisingly low tolerance for deviance in the abstract. Modal responses to psychological subscales (Blame, Anger, Anxiety and Depression) suggested many subjects frequently feel anger and anxiety, but little alienation or depression. The modal response denied having ever been sexually assaulted or beaten by an adult. The average respondent denied using any drugs except alcohol and marijuana in the previous month. The modal response was to report using both alcohol and marijuana 3-9 uses in the previous month.

Results indicated several significant differences for both gender and ethnicity. Overall, females reported worse family experiences, sexual abuse histories, depression and anxiety symptoms than males reported. In the assessment of accuracy of self-reported drug use, females appeared marginally more likely than males to be in the disagreement group (wherein self-reported marijuana use was in disagreement with negative UA results) than to be in the agreement group (wherein self report and UA results were either both positive or both negative). In contrast, ethnicity was not a significant variable in the assessment of accuracy of self-reported drug use. Sufficient differences were observed between Mexican-Americans and Spanish-Americans to justify keeping these two Latino groups separate in subsequent analyses. Within seven univariate analyses of significant multivariate effects for ethnicity, five analyses revealed WnHs reporting more problematic experiences and conditions than at least one other ethnic group. WnHs reported less family caring, family stability, religious values than either Latino subgroup, and more sexual abuse victimization than was reported by Spanish-Americans. WnHs, together with Mexican-Americans, reported more physical trauma/abuse than did Spanish-Americans. Two exceptions to this generalization were that Spanish-Americans reported more school deviance than the other two groups, and Spanish-Americans reported using more cocaine in the previous month than did WnHs. However, ethnicity never accounted for more than a small proportion of the variance in any of the variables examined, and does not appear to be a particularly useful variable in understanding or intervening in these variables.

These results are discussed in relationship to study hypotheses, including implications for a recently proposed model of a developmental pathway to delinquency for girls, and practical usefulness of these results is interpreted. Limitations of the study are discussed, as are implications for future research efforts.