The Significance of Power and the Power of Significance: Recommendations for Occupational Therapy Research
Research in the behavioral and social sciences including occupational therapy has been shown to be associated with low statistical power and a high rate of Type II experimental errors. Three methods of increasing power that are frequently suggested are increasing sample size, increasing effect size, and increasing the significance level The first two alternatives are often not possible in applied fields such as occupational therapy, and the third is generally not considered desirable since it leads to increased Type I error rates. A fourth alternative is proposed, which involves the partitioning of the decision region into three sections. This procedure is based on the Neyman and Pearson (1933) decision-theory approach to significance testing and is particularly applicable to areas of applied and clinical investigation such as occupational therapy. A sample power table is presented along with formulas to compute table values. The argument is made that using the procedures described will provide a method of unambiguously interpreting nonsignificant results and increase the power and sensitivity of occupational therapy research.