On ensuring the availability of evaluative data for secondary analysis

This paper is based on a programmatic effort to elicit and reanalyze data stemming from the evaluation of social programs. The stress is on identifying problems and negotiating for data and analyzing it. This includes: researchers' reluctance to disclose data for documentation and data control systems; vague or nonexistent policy for assuring access. The possible solutions presented here are based on governmental regulation and policy for publicly supported evaluation.Individuals who kindly provided their reactions to earlier drafts of this paper and/or advice and information in other ways include: Ole Engberg, John Evans, Harold B. King, Sheldon Laube, William Madow, Richard Marciano, Margaret E. Martin, Keith Marvin, Marjorie Powell, Alice Robbin, Peter Rossi, Lee Sechrest, Ernst W. Stromsdorfer, and Frederick Mosteller. Not all their reactions and criticisms have been accommodated, and we alone bear responsibility for the opinions registered here.Development of this paper has been supported by NIE Contract C-74-0115. All unpublished documentation discussed here is on file at the Methodology and Evaluation Research Division (Lucina Gallagher, Librarian).