Figure I.

Effect of Treatment on Readmission, Study 3 (Juvenile Detention)

Sample consists of the N = 2,693 youth admitted to the Cook County JTDC during period when random assignment was in effect, and for whom we have at least 18 months of follow-up data. Since randomization occurred by day of admission, day-of-admit fixed effects help control for any slight differences across days in treatment assignment probabilities. Graph shows effects conditional on baseline covariates as described in text, and with versus without day-of-admission fixed effects. Error bars represent 95% confidence intervals. CM is control mean; CCM indicates control complier mean.