Statistical Power Analysis: A Simple and General Model for Traditional and Modern Hypothesis Tests (repost)
Statistical Power Analysis: A Simple and General Model for Traditional and Modern Hypothesis Tests
Routledge; 2 edition | August 1, 2003 | ISBN-10: 0805845267 | 128 pages | PDF | 7.85 Mb
Routledge; 2 edition | August 1, 2003 | ISBN-10: 0805845267 | 128 pages | PDF | 7.85 Mb
This book presents a simple and general method for conducting statistical power analysis based on the widely used F statistic. The book illustrates how these analyses work and how they can be applied to problems of studying design, to evaluate others' research, and to choose the appropriate criterion for defining "statistically significant" outcomes. Statistical Power Analysis examines the four major applications of power analysis, concentrating on how to determine:
*the sample size needed to achieve desired levels of power;
*the level of power that is needed in a study;
*the size of effect that can be reliably detected by a study; and
*sensible criteria for statistical significance.