Association for Public Policy Analysis & Management

We report the results from a mixed-methods randomized controlled trial evaluation of Check & Connect, a mentoring program aimed at reducing truancy. Approximately 450 students in the Chicago Public Schools in grades 1-8 who had missed between two and six weeks of school during the previous school year were assigned a mentor for two years. Random assignment was carried out at the school, grade, and student level, allowing for experimental estimates of both participation in the program and spillover effects of the program on non-participating students. We find that program participation caused significant reductions in school absences among older students, but not among younger students. Findings from in-depth interviews of participating and control students and their families, and of mentors, are used to help understand variation in the effectiveness of the program to reduce truancy and increase student engagement.