Assessment of statistical power in contemporary strategy research

Abstract The concept of statistical power was reviewed, and the power of 44 recently published empirical studies in strategic management was analyzed. Using small, medium, and large estimates of effect size, standardized 0.05 α, and assuming nondirectional nulls, the mean power figures were 0.23, 0.59, and 0.83 for the three levels, respectively. These results were generally similar to findings in other social sciences and were considered particularly important for strategic management research, given the correlational nature of many strategic management investigations and the complexity and evolving stage of the field which make small effect size more likely. Ways to improve statistical power in strategic management research were discussed.