Grantee: Colorado Seminary. The full study report submitted to Arnold Ventures can be found here.
Description of the Intervention: Fostering Healthy Futures for Teens (FHF‑T) is a 9‑month intensive mentoring program for 8th and 9th grade students who have open child welfare cases due to maltreatment within their families. The program was adapted from the Fostering Healthy Futures program for pre-adolescent youth, for which two RCTs have found sizable impacts on certain child and young adult outcomes.
Study Design: The study sample comprised 245 teens (8th and 9th graders) in the Denver metro area with open child welfare cases due to maltreatment, recruited in four cohorts over four consecutive summers from 2015 – 2019. These youth were randomly assigned to a treatment group that participated in FHF‑T, or a control group that received child welfare services as usual.
The study’s primary, pre-specified outcomes were (i) juvenile justice involvement and (ii) self-reported delinquency. The first outcome was measured using court charge data from the State Court Administrator’s Office, with data collected through 15 months after program participation (25−28 months after study enrollment). The second outcome was assessed 1.5 years after program participation using youth-reported data collected via the Adolescent Risk Behavior Survey, which measures engagement in 16 different delinquent behaviors in the prior year.
Impact on the Primary Outcomes: Sizeable reductions were found for both pre-specified primary outcomes, but neither effect reached statistical significance. By the 25 – 28 month follow-up, the treatment group was (i) less likely to have had a court charge (27% in the control group vs. 13% in the treatment group, p=0.19), and (ii) less likely to self-report any delinquency in the past 12 months (65% in the control group vs. 57% in the treatment group, p=0.33).1 As discussed in the final report, during the study’s follow-up period, the advent of COVID-19 and changes in Colorado juvenile justice policies substantially reduced the rate of court charges for youth in Colorado. This resulted in fewer such charges for both the treatment and control group in this study, making it more difficult for the study to be able to detect an effect on charges.
Study Quality: This was a generally well-conducted RCT, but contextual factors and the smaller sample size made it difficult to reliably detect effects on crime outcomes.2
These are intent-to-treat effects — that is, the effects on youth who were assigned to the program. The authors also estimate treatment-on-treated effects (effects on youth who enrolled in the program) as an exploratory analysis in the full report.↩︎
For example, the study had successful random assignment (as evidenced by similar treatment and control groups at baseline), and valid analyses that were pre-registered. The study had no sample attrition in measuring court charges, and reasonably low and non-differential attrition in measuring self-reported delinquency (23% attrition in the treatment group vs. 19% in the control group).↩︎