Study 1
Pratt, Holsinger, and Latessa (2000) used a quasi-experimental design to evaluate the impact of the Turning Point Multiple DUI Treatment Program on recidivism measures at the 10-year follow-up period. The sample comprised the same participants from a previous 1-year (Langworthy and Latessa 1993) and 4-year follow-up evaluation (Langworthy and Latessa 1996). The treatment group (n = 531) consisted of individuals who met the program eligibility criteria and agreed to participate in the Turning Point program; the comparison group (n = 192) consisted of individuals with multiple DUI convictions who were adjudicated at the same time as those in the treatment group, but chose not to participate in the intervention.
Most program participants were male (90 percent) and white (77 percent), followed by African American or other race/ethnicity. The average age was 32.6 years, and most program participants (78 percent) had two to five prior DUI convictions. The average length of stay for Turning Point participants’ current incarceration was 50.7 days. Most of the comparison group (96 percent) was male and white (71 percent), followed by African American or other race/ethnicity. The average age of the comparison group was 40.7 years, and most (91 percent) had two to five prior DUI convictions. The average length of stay for the comparison groups’ current incarceration was 77.1 days. Significant differences were found between groups with respect to sex, age, employment status, number of children, number of prior DUI convictions, and amount of time spent in jail for the current offense. These variables were included in the regression analyses as statistical controls. No significant differences were found for measures of race, education, income, and marital status.
Using administrative data, the researchers examined three recidivism outcomes at the 10-year follow up: arrests for any new offense, arrests for an alcohol-related offense, and arrests for a DUI offense. The regression model controlled for six variables on which the groups were found to differ (listed above), as well as the amount of time spent at risk in the community during the follow-up period.
One limitation of the study was that members of the comparison group may have received other substance-use treatment services during the study period such as hospitalization, intensive supervision, or Alcoholics Anonymous groups. However, information was not collected on the extent to which the comparison group received these other services. Furthermore, the authors did not provide any information regarding program dropouts or non-completers. No subgroup analyses were conducted.