Whether in the laboratory or the field, one of the more important challenges for researchers, particularly those who study treatment effectiveness, is designing research that will have sufficient sensitivity to detect those effects it purports to investigate. The sensitivity of a research design is determined by many factors ranging from sample size, validity and reliability of the measures and the type of statistical analysis. However, the end result is a data set upon which the statistical significance can be tested using procedures that have a high level of statistical power. This book examines the concept of design sensitivity and explains statistical power and the elements that determine it. Through explanations of examples, Lipsey discusses a variety of topics, such as the many factors that degrade design sensitivity, effect size parameters and various approaches to assessing it and how to estimate statistical power for various statistical tests.
Mark W. Lipsey is the Director of the Center for Evaluation Research and Methodology, and a Senior Research Associate, at the Vanderbilt Institute for Public Policy Studies (Ph.D. in Psychology from The Johns Hopkins University in 1972). His professional interests are in the areas of public policy, program evaluation research, social intervention, field research methodology, and research synthesis (meta-analysis). The topics of his recent research have been risk and intervention for juvenile delinquency and substance use, early childhood education programs, and issues of methodological quality in program evaluation research. Professor Lipsey serves on the editorial boards of Evaluation and Program Planning, Psychological Bulletin, the Journal of Experimental Criminology, and the American Journal of Community Psychology, and boards or committees of, among others, the National Research Council, the Department of Education What Works Clearinghouse, Campbell Collaboration, and Blueprints for Violence Prevention. He is a recipient of the American Evaluation Association's Paul Lazarsfeld Award, the Society of Prevention Research's Nan Tobler Award, a Fellow of the American Psychological Society, and co-author of the program evaluation textbook, Evaluation: A Systematic Approach and the meta-analysis primer, Practical Meta-Analysis.
PART ONE: STATISTICAL POWER IN TREATMENT EFFECTIVENESS RESEARCH Treatment Effectiveness Research and Design Sensitivity The Statistical Power Framework Effect Size The Problematic Parameter How to Estimate Statistical Power PART TWO: USEFUL APPROACHES AND TECHNIQUES Dependent Measures Design, Sample Size, and Alpha The Independent Variable and the Role of Theory Putting It All Together
"A very useful book to social science researchers and a valuable reference for practicing statisticians. I found many good suggestions in it. This book is a thought-provoking essay on statistical power analysis. It is worthwhile reading." * Journal of the American Statistical Association` *
