Prediction of Absconding from a Halfway House Damon Mitchell, Central Connecticut State University Jake Villeneuve, The Connection, Inc Raymond Chip Tafrate, Central Connecticut State University Introduction Overview While predictors of recidivism and technical violations are often examined in probation and parole outcome research, predictors of absconding have received less attention. Williams, Mc. Shane, & Dolny (2000) found that parole absconding was best predicted by three variables: an unstable living arrangement, frequent unemployment, a prior parole violation. Schwaner (1997) found that compared to non-absconders, parole absconders: had a more extensive criminal history, had a more extensive history of parole violation/revocation, were more likely to abuse alcohol. Mayzer, Gray, & Maxwell (2004) found that probation absconders: were similar to probation revokees in their risk level and number of prior felonies, differed from probation revokees in that they had fewer prior misdemeanors and an older age of criminal onset, differed from those who successfully completed probation in that they were less likely to be employed, had more address changes, and had more supervision conditions. Unfortunately, the extent to which the prior research generalizes to parolee/work-release clients in a halfway house in unknown. These clients differ from those in the studies noted above in that they are living in a structured and supervised setting rather than independently in the community, and are having their day to activities overseen by a case manager rather than a supervision officer capable of imposing sanctions. The Present Study We examined the ability of demographic variables, criminal history variables, and assessments of criminal risk, substance abuse, and criminal thinking to predict the absconding of male offenders on parole or work-release admitted to a halfway house. Method Participants 136 male offenders on parole or work-release admitted to a halfway house. Average of 32 (SD = 9) 54% Black, 28% White, 18% Latino Average of 10 arrests (SD= 9), 3 incarcerations (SD = 3) Halfway House An all male, 30 bed, post incarceration facility located in an urban setting. The facility utilizes a case management service model to develop individualized service plans that address clients risks and needs. Services include: vocational counseling, housing/transitional Planning, cognitive skills and daily living skills training, substance Abuse education, GED classes, & AA/NA meetings. A typical course of stay is 90 to 120 days. Clients are assigned a case manager who works with them to complete their assessments and identify key risk/need factors. The intake process includes assessment of the client’s criminal history, substance abuse use, emotional stability, employment concerns, and other risks/needs. After the intake process has been completed, the clients complete five days of orientation during which time they are not allowed to leave the facility. After orientation, clients are allowed out of the facility to work, job search, address basic needs, and attend mental health substance abuse treatment, or church. Clients attend a job search group at least twice a week. While at the facility, clients are assigned to one or more group interventions addressing criminal and addictive thinking, relapse prevention, and community reentry challenges. When the clients are nearing their end of stay, their case manager will complete a discharge plan with the client. Assessment Level of Service Inventory-Revised (LSI-R; Andrews & Bonta, 1994): This interview based criminal risk/need instrument was administered to clients at intake. Adult Substance Use Survey-Revised (ASUS-R; Wanberg, 2005): This self-report measure of substance abuse, criminality, and mood was also administered at intake. The ASUS-R yields scores on 15 subscales, 9 of which were included in the present study. Criminogenic Thinking Profile (CTP; Mitchell & Tafrate, 2011): This self-report measure of criminal thinking was also administered to clients at intake. The CTP yields a Total Score as well as scores on 8 subscales, each reflecting a different thinking pattern (see Table 1). Table 1 1 Table Procedure The present study was conducted through analysis of an existing database that contained demographic, criminal history, intake assessment, program, and outcome data on all clients discharged from the halfway house over a 2. 5 year period (N = 146). Data from 10 clients was removed because they were transferred from the program due to medical or other reasons. Predictor Variables Demographic Age Highest level of education completed Race (coded as Black/Non. Black) Criminal history Number of arrests as an adult Number of incarcerations as an adult Number of tickets for institutional misconduct Amount of time served for instant offense Results Predictor Variables (continued) Table 1 Descriptive Data 13% of clients (n = 18) absconded from the halfway house. Average length of time from admission to absconding was 96 days (SD = 63 ). Logistic Regression Analyses The dependent variable, absconding, was dichotomously coded (0 = nonabsconder; 1 = absconder). 6 logistic regression analyses were conducted (demographics, criminal history, LSI Total Score, CTP subscales, ASUS-R subscales) and yielded the following findings: 1) Ethnic background (Black) was the only demographic variable that was a significant predictor of absconding (Wald = 7. 11, p <. 01, Odds Ratio = 8. 28), 2) No criminal history variables d emerged as significant predictors, 3) LSI Total Score was not a significant predictor, 4) Higher Total Score on the CTP was predictive of absconding (Wald = 3. 92, p <. 05, Odds Ratio = 1. 03, 5) Higher scores on the Disregard for Other and Poor Judgment subscales of the CTP were predictive of absconding (see Table 2), 6) No ASUS-R subscales scores emerged as significant predictors. Conclusions Absconding from the halfway house was more likely to occur among African American clients than among White or Hispanic clients. Table 2 Greater self-reported criminal thinking overall was associated with absconding, and in particular, thinking patterns that devalue others and underestimate the negative outcomes of risky behavior. Neither criminal history nor substance abuse, two established predictors of recidivism, were predictive of absconding from the halfway house. A commonly employed risk assessment tool, the LSI-R, did not predict absconding. The results suggest that risk of absconding from a halfway house may be more associated with the thought patterns that underlie antisocial behavior than established behavioral predictors of recidivism.
Скачать презентацию Prediction of Absconding from a Halfway House Damon
e779be042c2fa42a14fb4cadcbb98b29.ppt