A program evaluation by Hall and colleagues (2004) used a prospective, longitudinal design to look at the outcomes among 214 women who either participated in the Forever Free program or participated in a comparison condition. The treatment group consisted of 119 female clients entering the program between October 1997 and June 1998 who volunteered to participate in the study. The comparison group consisted of 95 women enrolled in Life Plan for Recovery, an 8-week substance abuse education course, between April and November 1998. There were no statistically significant differences between the treatment and comparison groups at study intake on most measures. Women in both groups were about 35 years of age. They averaged about 16 prior arrests and 8 prior incarcerations, and most were incarcerated for a drug offense. The sample was approximately 34 percent African American, 33 percent White, 21 percent Latina, and 7 percent Other.
The main outcome variables included crime or recidivism rates, drug use, and employment. Measures were taken from study participants at intake and follow-up, approximately 1 year after women were released from prison (between September 1999 and August 2000). For the follow-up interviews, 101 interviews were completed with Forever Free program participants (85 percent of the original sample) and 79 interviews were completed with the comparison group (82 percent of the original sample).
Information was collected on subjects’ criminality, including age at first arrest, lifetime arrests, most serious charge for current case, arrests and incarcerations since release, and incarceration status. Information was also collected on subjects’ substance abuse history and treatment for the periods before intake and since release, including primary substance of abuse, age of first use, age of first regular use, day of use in the 30 days before incarceration, use since release, use in the 30 days before the follow-up interview, use per day, and street value per day for the following drugs: alcohol, amphetamines/methamphetamines, barbiturates, cocaine (crack), cocaine (powder), heroin and other opiates, marijuana, PCP, tranquilizers, and other drugs. The California Department of Corrections Offender-Based Information System provided additional information on subjects’ incarceration history, including date and charge of each incarceration, parole release date, parole region and status, and date and reason for return to custody (new charge or parole violation).
The study used bivariate analyses, including chi-square analysis and t-tests, to determine program effects on outcome measures. In addition, multivariate analyses, including logistic regression and Cox regression, were used to find any predictors of long-term success while controlling for background characteristics.