Study
The quasi-experimental study by Alarid and Rangel (2018) focused on juveniles on probation who participated in Operation Night Light (ONL) in a midsize city in the Midwest. The total sample comprised 724 youth: there were 287 ONL probationers matched with a comparison group of 437 regular persons on probation. The comparison group was generated from a list of youth on regular probation who met ONL criteria and likely would have been assigned to ONL if the program had been available at the time. The treatment group received home visits from a probation officer and a police officer about once every other month, for an average of five visits total. The comparison group received no home visits. Both groups were required to attend probation office visits and adhere to treatment plans. None of the youth were under active supervision at the time of the study, but all were tracked for 24 months following the end of their probation.
The ONL group was 46.7 percent white or Hispanic, 53.3 percent African American, and 79 percent male. ONL youth were on probation for person/violence offenses (22.7 percent), property offenses (45.5 percent), drug possession (15.0 percent), and for other offenses (16.8 percent). Fifty percent were adjudicated for a felony offense, and 50 percent were adjudicated for a misdemeanor. On average, ONL youth were on probation for 11.3 months. The comparison group was 43.7 percent white or Hispanic and 56.3 percent African American, and 84.7 percent were male. Comparison youth were on probation for person/violence offenses (39.1 percent), property offenses (38 percent), drug possession (20.1 percent), and for other offenses (2.7 percent). Of this group, 74.6 percent were adjudicated for a felony offense, whereas 25.4 percent were adjudicated for a misdemeanor. On average, comparison youth were on probation for 10.4 months. In both groups, ages ranged from 10 to 18 years with an average of 15. Participants who were 18 at the time of data collection had been placed on probation prior to their 18th birthday.
Chi-square analyses were used to uncover any potential statistically significant differences between the two groups. The ONL group and comparison group were found to be similar in terms of sex, race, ethnicity, age, proportion of drug and nonviolent offenses, and length of time on probation. The only statistically significant difference between the two groups was the proportion of person/violent crimes. The comparison group had more person/violent offenses (39 percent) than the ONL group (23 percent). The comparison group also had a higher proportion of felony offenses (75 percent) than the ONL group (50 percent).
Official agency data for descriptive variables was obtained from juvenile probation case files (i.e., age, race, gender, current offense, offense severity). Administrative case files and probation records provided the data for number of home visits, type of home visits (no response, contact with the person on probation or collateral contact), whether a new crime was committed while on probation, disposition of the new crime, and type of discharge from probation supervision (successful completion, revocation for a new crime, or revocation for a technical violation). During the study, there were a total of 1,420 home visit attempts made to the ONL group, which resulted in an average of five visits per youth.
One outcome of interest was defined as successful or unsuccessful completion of probation. Successful completion of probation occurred when a youth was discharged from probation after serving the time. Unsuccessful completion was coded as revocation, which occurred when a youth was terminated or discharged prematurely from probation due to severity of a new crime. Severity was determined by whether the new offense was a misdemeanor or a felony. Those on probation who committed technical violations or a crime while on probation, but remained on supervision, were not considered discharged until probation ended. Discharge from probation was counted only once per youth. Outcome variables included prevalence of rearrest, type of crimes involved in the rearrest, and date of rearrest in relation to date of probation discharge.
Logistic regression models were run with both treatment and control groups together to determine which variables were significant predictors of probation completion and later rearrest. Survival analysis and Cox regression were used to analyze factors involved in a youth’s time to rearrest. The standard Cox regression analysis was used to examine the time until individuals in either group would recidivate after being discharged from probation.