This readable and comprehensive text is designed to equip students and practitioners with the statistical skills needed to meet government standards regarding public program evaluation. Even those with little statistical training will find the explanations clear, with many illustrative examples, case studies, and applications.
Far more than a cookbook of statistical techniques, the book begins with chapters on the overall context for successful program evaluations, and carefully explains statistical methods--and threats to internal and statistical validity--that correspond to each evaluation design. Laura Langbein then presents a variety of methods for program analysis, and advise readers on how to select the mix of methods most appropriate for the issues they deal with-- always balancing methodology with the need for generality, the size of the evaluator's budget, the availability of data, and the need for quick results.
1. What This Book Is About
2. Defensible Program Evaluations: Four Types of Validity
3. Internal Validity
4. Randomized Field Experiments
5. The Quasi Experiment
6. The Nonexperimental Design: Variations on the Multiple Regression Theme
7. Designing Useful Surveys for Evaluation
8. Summing It Up: Meta-Analysis