Study 1
Blakeslee and Keller (2018) examined the effects of My Life Mentoring by investigating outcomes of a combined sample of participants from two large randomized controlled trials that used parallel measures and designs (Geenen et al. 2013; Powers et al. 2012). In both trials, program participants were compared with control youth who received regular transitional services. The interventions took place in Multnomah County and Clackamas County, Oregon, which include urban areas (i.e., Portland), suburban areas, and rural areas. Eligible participants (n = 293) were between 16.5 and 18.5 years old, had been in foster care under the guardianship of the Oregon Department of Human Services for at least 90 days, and lived within the target area of the study. Youth who were in crisis were excluded, as were youth who were moving out of state, youth who were non-English speaking, and youth who could not access the community as needed for the intervention due to incarceration or placement in a locked residential facility. Treatment youth (n = 144) received weekly 1-hour, one-to-one mentoring and access to frequent peer-led group mentoring workshops in addition to regular transitional services. Control youth (n = 149) received regular transitional services. These services typically included the Independent Living Program, which consisted of classes to prepare for independent living, and one-on-one work with a case manager to develop a personal transition plan.
The average age of participants in the treatment group was 17.3 years, and 57.6 percent were female. Of this group, 43.8 percent were white, 24.2 percent were classified as Native American, 18.0 percent were Hispanic, 16.7 percent were African American, 2.0 percent were Asian, 2.0 percent were Pacific Islander, and 31.8 percent were multiracial/other. In terms of residence, about 60.0 percent of treatment youth lived in non-relative foster care, 29.2 percent lived in relative foster care (kinship), 4.9 percent lived in group home/residential treatment, and 3.3 percent lived elsewhere (with a friend/in own apartment). Participants had been in foster care for an average of 6.0 years with 41.7 percent of youth experiencing a change in placement in the past year. About 96.5 percent were attending school or in a GED program, 60.0 percent had a special education disability, 39.6 percent were enrolled in the Independent Living Program, and 23.6 percent were receiving developmental disability services.
The average age of participants in the control group was 17.3 years, and 49.7 percent were female. Of this group, approximately 48.3 percent were white, 18.1 percent were African American, 17.4 percent were Hispanic, 8.7 percent were classified as Native American, 1.3 percent were Asian, 0 percent were Pacific Islander, and 23.5 percent were multiracial/other. About 68.0 percent lived in non-relative foster care, 22.8 percent lived in relative foster care (kinship), 5.4 percent lived in group home/residential treatment, and 4.0 percent lived elsewhere (with a friend/in own apartment). Participants had been in foster care for an average of 6.0 years with 35.0 percent of youth experiencing a change in placement in the past year. Of the control youth, 90.0 percent were attending school or in a GED program, 59.1 percent had a special education disability, 44.3 percent were enrolled in the Independent Living Program, and 22.1 percent were receiving developmental disability services.
Youth were blocked for group assignment based on participation in the Independent Living Program and whether they received special education or disability services. There were no significant differences in race/ethnicity, special education status, developmental disability status, gender, or baseline delinquency scores between groups who were followed up at the final time point (treatment n = 77; control n = 81).
Self-report measures were administered to measure self-determination (The Arc’s Self-Determination Scale; Wehmeyer and Kelchner 1996) and self-efficacy (My Life Activity Checklist; Blakeslee and Keller 2018) at multiple time points: baseline (Time 1), at the end of the intervention year (Time 2), at 1-year post-intervention (Time 3), and at 2-years post-intervention (Time 4). Transition outcomes, including high school status, homelessness, and past-year trouble with the law, were measured in a self-report survey at Times 2 through 4. Criminal justice outcomes were measured at Times 1 through 4. This included charge severity across six types of charges, calculated by assigning each of six categories (i.e., property crime, drug-related crime, assault, robbery, sexual offense, other) a score of 1–3 (0 = no arrest, 1 = arrest only, 2 = misdemeanor conviction, 3 = felony conviction) and summing those scores for a final score of 0–18. Range of punitive system involvement was measured as the sum of past-year days in jail, in prison, or on probation and/or parole. Comprehensive criminal justice involvement was a yes/no question on whether the participant had any past-year trouble with the law or juvenile justice, and/or any self-reported arrests or convictions, and/or any self-reported days incarcerated or in trouble with the law. Two self-report measures of delinquency were also included at the final time point. Findings were assessed using general linear mixed models controlling for baseline age and gender. Subgroup analysis was conducted on implementation scores (coaching mode, skill building focus, or total coaching hours).
