Statistics for the hand surgeon. Part 2: avoiding common pitfalls.

Abstract

In this second of a two-part article, we describe some of the common statistical pitfalls encountered in hand surgery research. These include dichotomania, the 'Table 2 fallacy', p-hacking, regression to the mean, overfitting and unaccounted data clustering. We explain the impact of these pitfalls on hand surgery research and describe techniques to avoid them. The aim of this two-part article was to provide a starting point for hand surgeons to refer to when conducting or analysing research and provide resources and references for interested readers to explore.