The Effect of Mentoring on School Attendance and Academic Outcomes: A Randomized Evaluation of the Check & Connect Program

large-scale RCT in Chicago Public Schools, grades 1-8

At-risk students in Chicago Public Schools, grades 1-8

Large randomised controlled trial of Check & Connect, the most widely endorsed intensive attendance/engagement mentoring programme in US practice, run in partnership with the district. Measures attendance, test scores, GPA, and within-school spillovers to non-participating students. This is the strongest test available of whether an intensive, high-cost attendance intervention converts attendance into learning.

Attendance moved substantially in grades 5-7: 4.2 fewer days absent, a 22.9% reduction. No detectable effects at all in grades 1-4. And there were NO statistically significant effects on learning outcomes - neither test scores nor GPA - and no detectable spillovers to other students in treated schools. The authors themselves note the modest impact per dollar relative to both cheap nudges and more intensive interventions.

Low - randomised, large, district-partnered. This trial is decisive for the second step of the causal chain the topic evaluates. A 22.9% reduction in absences is a very large behavioural effect by the standards of this literature - larger than any of the messaging trials - and it produced nothing measurable on achievement. If the observational absence-achievement gradient reflected a real instructional dose-response, this trial should have detected it.