Purpose: Knowledge of the risk and protective factors associated with adolescent problem behaviors such as delinquency, drug use, or school failure has increased significantly. Most evidence about adolescent risk and protection has been derived from general youth populations or from study samples in juvenile justice or other systems of care. Relatively little is known about specific patterns of risk and protection among children and adolescents raised in conditions of poverty. Young people from economically disadvantaged backgrounds are often characterized by their similarities in relation to exposure to crime and violence, school dropout rates, and barriers to economic success. Such general characterizations mask important individual differences and fail to adequately describe underlying patterns of risk, protection, and antisocial behavior that are present in the lives of children and youth who live in poverty. To better understand such patterns, we investigated risk and protective factors in a sample of 150 youth between 6th and 8th grade who resided in three urban public housing developments. The relationship between identified patterns of risk and protection and educational and behavioral outcomes was also assessed.

Methods: A self-report survey was conducted with 150 youth residing in three urban public housing communities. Youth were surveyed in April 2005; sixty-six percent (N= 99) were Latino/a, 14% (N=21) were African-American, 13% (N=20) were Asian-American, and 7% (N=10) were Caucasian or mixed ethnicity. Measures of risk and protection included individual characteristics, peer relationships, family conditions, and neighborhood-level variables. Individual and peer factors were assessed using the Rosenberg Self-Esteem Scale, the Morgan-Jinks Student Efficacy Scale, the Youth Coping Index, the Index of Peer Relations, and additional self-report items that evaluated subjects' attitudes toward antisocial behavior. Family level measures included five subscales of family bonding. Neighborhood items were evaluated by the Daily Hassles Questionnaire and additional items that measured subjects' perceptions of neighborhood and community. Educational and behavioral outcome measures included level of participation in the local after-school program, grades and standardized test scores, and teacher reports on the Carolina Child Checklist. Self-reports of drug use, delinquency, and other antisocial behavior were obtained. Cluster analytic techniques were used to identify patterns of risk and protection. Cluster validation included internal comparisons with variables in the model and external comparisons with those variables not included in cluster groupings. Analysis of variance was used to assess the relationship between cluster groupings and external variables and regression analyses were conducted to evaluate the relationship between cluster groupings and outcome variables.

Results and Implications for Practice: K-means clustering techniques revealed three cluster profiles characterized by unique patterns of risk and protection. External comparisons yielded significant differences in educational outcomes and teacher reports by cluster. Cluster membership was significantly related to a scaled index of antisocial conduct. Study results suggest that there are important differences in patterns of risk and protection among young people raised in poverty. Implications for the development of interventions that account for these differences are discussed.