Wayne N. Welsh

The Effects of School Climate on School Disorder

Explanations of school disorder have suffered from at least two deficits: (1) institutional explanations of disorder (that is, school climate) have been largely ignored, and (2) insufficient attention to appropriate measures of disorder has guided research and policy. Like people, schools have their own characteristic personalities, or climates. Using survey responses from students in middle schools in Philadelphia, the author discusses the effects of school climate (such as clarity and fairness of rules) and individual student characteristics (such as age, sex, race, and dimensions of bonding) on different measures of school disorder, including victimization, avoidance, perceptions of safety, misconduct, and offending. The schools varied significantly on all measures of disorder, and school climate provided significant explanatory power for each. Results varied for different measures, though. For example, school climate predicted less serious misconduct more strongly than it predicted serious offending. School climate offers significant potential for enhancing both the understanding and the prevention of school violence.

A Macro-Level Model of School Disorder:

Drawing on school climate theory and social disorganization theory, this article examines the influence of major institutional and community factors on disorder in Philadelphia public schools. Using U.S. census data, school district data, and police data, the authors examined the following predictors of disorder in 43 middle schools: community poverty and residential stability, community crime, school size, and school stability. Community was conceptualized in two ways: local (the census tract around the school) and imported (aggregated measures from the census tracts in which students actually reside). Previous studies have failed to make this distinction when assessing community-level influences on school disorder. The authors used path analysis to examine direct and indirect relationships between community characteristics (poverty, residential stability, crime rates), school size, school stability (a factor score based on student attendance and turnover), and school disorder (a factor score based on school incident data and dismissal rates). The local community model fit the data better than the imported model: The communities immediately surrounding schools have a stronger influence on school disorder than the communities from which students are drawn. Community poverty exerted strong indirect effects on school disorder in both models. The effects of community variables on school disorder were strongly mediated by school stability, illustrating that analyses of institutional processes have much to add to the explanation of school disorder.

A Multisite Evaluation of Prison-Based Therapeutic Community Drug Treatment:

A quasi-experimental study examined multiple postrelease outcomes up to 2 years for inmates who participated in therapeutic community (TC) drug treatment programs (n = 217) or comparison groups (n = 491) at five state prisons. Statistical controls included level of need for treatment, current and prior criminal history, and postrelease employment. Prison TC was effective even without mandatory community aftercare, although main effects and interactions varied somewhat across different outcome measures and sites. TC significantly reduced rearrest and reincarceration rates but not drug relapse rates. Postrelease employment predicted drug relapse and reincarceration, and employment interacted with age to predict rearrest. Two sites had higher drug relapse rates than the other three. Implications for research and policy are discussed.

PREDICTING RECIDIVISM FOR RELEASED STATE PRISON OFFENDERS: Examining the Influence of Individual and Neighborhood Characteristics and Spatial Contagion on the Likelihood of Reincarceration

We examined the influence of individual and neighborhood characteristics and spatial contagion in predicting reincarceration on a sample of 5,354 released Pennsylvania state prisoners. Independent variables included demographic characteristics, offense type, drug involvement, various neighborhood variables (e.g., concentrated disadvantage, residential mobility), and spatial contagion (i.e., proximity to others who become reincarcerated). Using geographic information systems (GIS) and logistic regression modeling, our results showed that the likelihood of reincarceration was increased with male gender, drug involvement, offense type, and living in areas with high rates of recidivism. Older offenders and those convicted of violent or drug offenses were less likely to be reincarcerated. For violent offenders, drug involvement, age, and spatial contagion were particular risk factors for reincarceration. None of the neighborhood environment variables were associated with increased risk of reincarceration. Reentry programs need to particularly address substance abuse issues of ex-offenders as well as take into consideration their residential locations.

Cognitive mediators of aggression. Test of a causal model

Although theoretical and empirical evidence suggests that cognitions play a significant role in shaping aggressive responses, causal relations between cognitions and other predictors of aggression (e.g., past behavior, arousal, personality traits) are unclear. This study attempted to predict aggression in an experimental setting, using the Ajzen-Fishbein Theory of Reasoned Action. Fifty-one patients in a federal forensic hospital role-played aggressive behavior. Independent measures included cognitive components of the Ajzen-Fishbein model (attitude toward the behavior, subjective norm, and intention); personality measures of aggressiveness; self-report measures of anger and arousal; and indices of past behavior collected from archival data. Results from path analysis and multiple regressions suggested that (a) cognitive variables only partially mediated aggressive behavior; (b) arousal and trait anger exerted significant effects on aggression, independent of the cognitive components of the Ajzen-Fishbein model. Implications for theory and research are discussed.

The Predictive Validity of the Level of Service Inventory—Revised for Drug-Involved Offenders

The Level of Service Inventory—Revised (LSI-R) has been validated for a variety of criminal justice populations, although one large segment for which the instrument has not been validated is drug-involved offenders. With the increasing implementation of the LSI-R across correctional jurisdictions in the United States and internationally, and with a significant number of offenders having substance abuse histories, validation of the instrument for this population is overdue. Using bivariate correlations and logistic regression analyses, the LSI-R total score was found to be a stable, significant predictor of reincarceration for a sample of 276 drug-involved offenders. This finding held up in multivariate analyses even when controlling for a variety of other factors including age, criminal history, time-at-risk in the community following release from prison, and treatment-related variables.

Predictors of therapeutic engagement in prison-based drug treatment.

Few studies to date have examined predictors of therapeutic engagement (TE) or other indicators of responsiveness to prison drug treatment. Subjects were 347 inmates participating in a 12-month modified therapeutic community (TC) drug treatment program at a specialized treatment prison for convicted, drug-involved offenders. Data were obtained through correctional databases and the administration of the TCU Drug Screen II, the Resident Evaluation of Self and Treatment (REST), and the Counselor Rating of Client (CRC) form. Three main hypotheses were supported: (1) baseline motivation predicted therapeutic engagement net of other inmate characteristics; (2) critical dimensions of the treatment experience (e.g., peer support, counselor rapport) also predicted therapeutic engagement; and (3) dynamic predictors and programmatic characteristics became more important over time. Implications for research, theory and policy are discussed.

A Census of Prison-Based Drug Treatment Programs: Implications for Programming, Policy, and Evaluation

Despite a growing realization that unmeasured programmatic differences influence prison-based drug treatment effectiveness, few attempts to systematically measure such differences have been made. To improve program planning and evaluation in this area, we developed a census instrument to collect descriptive information about 118 prisonbased drug treatment programs in 24 state prisons. The census assessed program content and structure (e.g., program type, duration, and primary treatment approach), program staff (e.g., duties and staffing ratios), and inmates (e.g., eligibility and intake procedures). Collection of descriptive program information can greatly facilitate drug treatment program and policy planning, and meaningfully inform the design of subsequent outcome evaluations.

Reducing Minority Overrepresentation in Juvenile Justice: Results of Community-Based Delinquency Prevention in Harrisburg:

To reduce minority overrepresentation in its juvenile justice system, Pennsylvania sponsored nine community-based delinquency prevention programs in Harrisburg and Philadelphia. The authors describe results of their evaluation of the Harrisburg programs over a three-year period (1992 to 1995), reporting two-year follow-up data for the 1992-93 cohort (N = 191). The major objective was to reduce rates of arrest and rearrest for clients; programs also attempted to reduce major risk factors such as educational failure, dropout, and truancy. Using quasi-experimental comparison groups based on program attendance (control, low, and high), program effects are examined using logistic regression and multivariate analysis of variance. Programs were successful in reducing recidivism: over one-year and two-year follow-up periods, 2 recidivism was lowest for the high-attendance group. Program effects on school outcomes, however, were generally weak. Implications for policy and program development are discussed.

Effects of Prison Drug Treatment On Inmate Misconduct: A Repeated Measures Analysis

A small body of research supports the “treatment hypothesis” that participation in prison treatment programs reduces inmate misconduct, although methodological weaknesses have limited generalizable conclusions. Using general linear modeling repeated measures techniques, this study examined pre- and posttreatment misconduct for 1,073 inmates who participated in therapeutic community (TC) drug treatment (n = 294) or a comparison group (n = 779) at five state prisons. Predictors included age, length of sentence, drug dependency, and prior and current criminal history. The hypothesis that TC treatment alone would significantly reduce misconduct over time was not supported. Instead, changes in misconduct over time interacted with individual characteristics and time served posttreatment. The article discusses implications of these results for treatment policies and future research.