Statistics in medical research: Common mistakes

The misuse of statistics in medical studies has been discussed extensively with the conclusion that it is both unethical and can have serious clinical consequences. These errors can contribute to incorrect conclusions, compromise the validity of studies, and overestimate or underestimate the effects of treatment. To avoid making these errors, it is critical to consider their presence and understand statistical concepts. This practice will ultimately lead to the use of appropriate statistical techniques for specific research questions and the calculation of an appropriate sample size to guarantee adequate statistical power. Common statistical errors in medical research include sampling bias, the incorrect determination of sample, failing to adjust for multiple comparisons, misinterpreting p-values as a measure of effect size or clinical relevance, choosing incorrect tests for a particular data set, type I and II errors, data fishing, and publication bias. It is important that researchers interpret their results using appropriate statistical concepts by soliciting feedback from specialist statisticians.