Controversies in the Evaluation of Compensatory Education

This document answers the question, "What has been learned about evaluation methodology from the decade of compensatory education"? Ten issues dealing with the evaluation of Title I were identified within a general theoretical framework of evaluation. For each issue it was the aim of this document to do the following: 1)to clarify the issue, 2)to point out examples in which it is crucial, 3)to present and evaluate arguments on different sides of the issue, and 4)to suggest resolutions of the issue. Each of the issues was selected because its resolution is a necessary step in the development of a rational Title I evaluation policy. The issues addressed are: 1)To what should Title I treatments be compared? 2)Is longitudinal evaluation necessary? 3)When is representative sampling important? 4)How large a sample is necessary? 5)What constructs should be measured to determine Title I impact? 6)What types of achievement measurement instruments should be used in Title I evaluation? 7)What units of measurement should be used? 8)What are the conditions for valid comparisons between nonequivalent treatment and comparison groups? 9)Under what conditions can relationships of Title I costs and treatments to effectiveness be inferred? 10)How should data be aggregated across projects in Title I evaluations? (Author/AM)