Principles of Statistics for Surgeons III: Factors Influencing Sample Size Estimation

Abstract

The present article, the third in a series, describes some important ideas underlying the practice of sample size estimation. These include types I and II errors, the effect size, and data variability. Each of these ideas is described and illustrated in some detail without too much technical development, and should be relatively easy to understand for most surgeons. In particular, important concepts in elementary statistics are highlighted and should be useful for self-study. The necessity of some prior context-specific knowledge, when estimating the sample size, is emphasized. Simple numerical examples are presented at the end of the article to strengthen the emphasis. We hope the reader will be sufficiently informed in preparation for consultations with statisticians when such needs arise.