As an RPP, CAFÉCS has developed distributed leadership practices that enable the melding of cognition and design into the social context of the Chicago Public Schools (CPS). One such leadership practice is the CAFÉCS improvement cycle. Through weekly meetings of the CAFÉCS district and research leadership team and monthly meetings of the whole CAÉFCS district and research team, a problem of practice emerged around anecdotal reports of a high course failure rate in the first year of the implementation of a new computer science graduation policy. Given prior research in CPS that failure of one core class can significantly reduce the probability of graduating, this problem of practice threatened to derail the initiative. Typical of RPPs with high levels of trust, the whole team brainstormed hypotheses about potential factors affecting course failure, including student, teacher, and school characteristics. The analyses of these hypotheses revealed that teacher attendance at the ECS professional development had a significant impact on reducing failure. By first engaging in research to define the problem from a learning sciences perspective, the results had an immediate impact on practice. District implementation staff put the results into practice to significantly increase motivation and attendance at the ECS workshops in the next year. In addition, the results fed directly into design work on a hybrid course to address the social context of credit recovery for those students who do fail the class.