Statistical Power Analysis With Missing Data: A Structural Equa- tion Modeling Approach.

摘要

The need for power analysis, or, more specifically, sample size determination, cannot be overestimated. As Cohen (1988) helped applied researchers realize-particularly researchers in the social and behavioral sciences-studies often are either wastefully overpowered or, much more commonly, tremendously underpowered so as to be predestined to find few if any results. Thus, attention to the methods necessary for planning research is generally quite welcome. This is especially true given the increased complexity of the methods used in applied research, for commensurate with increases in methodological complexity come increased challenges in determining the required sample size to use those methods sensibly. For these reasons-furthering awareness of the importance of power as well as the facilitation of power analysis for complex models-this book is a positive addition to a still relatively small body of resources on the subject.