Multi-level models: Regression fixed effects.

In response to a need to implement an evidence-based prevention program, D.A.R.E. America adopted keepin’ it REAL. The program was previously developed and tested in middle school settings. As part of its adoption, an elementary version of the program was developed. This study tests the effectiveness of keepin’ it REAL when delivered to fifth graders. The intervention was delivered to two cohorts of students, the first in the 2019–2020 school year, the second in the 2020–2021 school year. Pretest surveys were completed by 6,122 students. The COVID-19 pandemic interfered with posttest and follow-up data collection. At immediate posttest, 2,049 students (33.5%) completed analyzable posttest surveys. One year after the pretest, 1,486 (24.3%) students completed usable follow-up surveys. We used algorithmically generated cases (virtual controls) that use treatment cases’ pretest psychosocial scores to assess program effectiveness. When compared to virtual control cases, the program had identifiable improvements in both a key psychosocial measure and in terms of deterring the onset of 30-day alcohol use, drunkenness, and vaping. Outcomes suggest that the delivery of elementary school keepin’ it REAL by D.A.R.E. officers is having a positive effect in terms of deterring the onset of alcohol use and vaping.