How Many Are Enough? Statistical Power Analysis and Sample Size Estimation in Clinical Research

One of the most common questions asked of statisticians is, “How many are enough?” The answer to that question, unfortunately, is complex, and only part of it can be answered by the statistician. This is because power calculation is part mathematical calculation and part judgment call. The former requires knowledge about population characteristics; the latter requires choices about tolerable error and meaningful effect size. The two come together to assure that the sample is sufficient to detect a phenomenon of interest and provide the reader with confidence in the results.