Abstract

This article refers to random error and how it may affect results of clinical trials, which is reflected in the confidence intervals that different authors calculate. In order to counteract random error a simple principle is applied – correct estimation of the sample size during the study design phase. In order to achieve this, the article describes and expounds on concepts such as interference and the factors that may lead to greater likelihood of random error occurrence in the results: variability, sample size and magnitude of differences. The article also mentions statistical significance and p value.