This report summarizes major findings and policy implications of a University of Pittsburgh project titled Methodological Research on Knowledge Use and School Improvement. The major emphasis has been methodological rather than substantive. The primary purpose has been to describe, evaluate, and recommend alternative concepts, methods, and techniques that enlarge capacities to understand and shape processes of knowledge use and school improvement. This project focuses on three core methodological problems: the criterion problem in knowledge use, the multiattribute problem of meanings attached to knowledge, and conceptualization and measurement in the knowledge transactional problem. Procedural inventory, typological procedures and analytic procedures are discussed as research objectives. The approach employed to realize these objectives may be described as a form of methodological triangulation: documentary and archival search; pooling, selection, and analysis of procedures; and network consultation, conferences, and workshops. The project results are summarized around five sets of methodological findings: reproducibility, reliability, and validity; knowledge use as a criterion; formal classification schema; generative classification schema; and sociocognitive grid procedures. Appendices contain an annotated bibliography of working and occasional papers, a sample procedure abstract, a sample study profile, and frequency distributions for study profiles. (Author/PN)