This study on accountability designs for underperforming schools focuses on probation as a tool for policy. Findings are based on the analysis of state performance data, the reading of approximately 100 school improvement plans, and the study of 11 focal schools in Maryland and Kentucky. Test-score data from Maryland show that probation is associated with an initial reversal of decline in the worst performing schools that fades out in subsequent years. Teachers in the 11 schools fundamentally do not buy into the accountability system. Low job commitment diminishes the threat of low-performance penalties. Many teachers were skeptical about the goals, felt unfairly judged by the system, and did not judge themselves according to the standards of the system. The accountability system also does not provide meaningful tools for self-evaluation. To a large degree, mild pressures and directives molded the teachers' responses to probation, with the principal being the system's conduit. The result was that probation, by itself, did not trigger much self-directed action on the part of the great majority of teachers. The four volumes and appendix contain the following parts: (1) analytical framework and study design; (2) the accountability systems of Maryland and Kentucky; (3) probation and performance motivation; (4) organization responses to probation; (5) planning for school improvement; (6) probation and instructional change; (7) summary of findings and discussion; (8) case studies; (9) the politics of accountability for school improvement in Kentucky and Maryland; (10) case study research guide; and (11) interview protocols. (RT)