Designing Small Evaluation Studies

"The book will be an important addition to instruction in designs for causal inference in the field of education. It is long overdue." - Thomas J. Lipscomb, The University of Southern Mississippi This text describes how to design and analyze small efficacy or evaluation studies, typically carried out as part of the development of programs or interventions in areas such as education. The problem facing many researchers is how to design a study that is as small as possible, yet big enough to yield relatively unambiguous evidence about an intervention's average effect. This text begins with an overview of validity, causal inference, statistics, effect sizes, and measurement. The authors then focus on designs for small, randomized trials, followed by a section on non-randomized causal designs: here they focus on three designs most useful for small studies including the non-equivalent control group, difference-in-difference, and interrupted time series designs. The final section summarizes the book, compares designs, discusses approaches to choosing a design, and provides guidance on reporting. Five case examples are used throughout the book to illustrate the material and there is a glossary of terms and concepts.