PERSONNEL PSYCHOLOGY zyxwvuts 1985. 38 STATISTICAL POWER OF TRAINING EVALUATION DESIGNS RICHARD D. ARVEY, DAVID A. COLE, JOY FISHER HAZUCHA, FRANS M. HARTANTO University zyxwvu of Minnesota Sample size requirements needed to achieve various levels of statisti- cal power using posttest-only, gain-score, and analysis of covariance designs in evaluating training interventions have been developed. Re- sults are presented which indicate that the power to detect true ef- fects differs according to the type of design, the correlation between the pre- and posttest, and the size of the effect due to the training program. We show that the type of design and correlations between the pre- and posttest complexly determine the power curve. Finally, an estimate of typical sample sizes used in training evaluation design has been determined and reviewed to determine the power of the various designs to detect true effects, given this sample-size specifi- cation. Recommendations for type of design are provided based on sample size and projected correlations between pre- and posttest scores. The power of a statistical test refers to the probability of correctly rejecting the null hypothesis when it is false. For example, if the null hypothesis asserts that there is no effect due to some kind of training intervention, power has to do with correctly accepting the alternative hypothesis that a real effect did, in fact, result. While industrial/orga- nizational psychologists have been familiar with the concept of statis- tical power for years, several recent publications have called attention to the idea that, in many instances, typical sample size may be inade- quate for detecting true group differences. In 1976, Schmidt, Hunter, and Urry demonstrated that the sample sizes necessary to detect significant validity coefficients in establishing predictor-criterion relationships were far greater than had previously been realized, especially when these variables had less than perfect reliability or restriction of range occurred on the predictor. For exam- ple, industrial psychologists were somewhat shocked to learn that the Support for this paper was provided by the School zyxw of Management, University of Minnesota. Dr. Scott E. Maxwell provided statistical consultation regarding the statisti- cal derivations found within the manuscript. COPYRIGHT zyxwvut 0 1985 PERSONNEL PSYCHOLOGY. INC 493