The calculation of the sample size is carried out while the research is still in the design phase. In research, multiple variables of different types are often examined with separate statistical tests. In this case, it is recommended to perform power analyses in accordance with the type of measurement of each variable and use the largest sample size obtained. If the research is an observational study, approximately 20% larger sample size is needed than in clinical trials. This is because in non-randomized studies, confounders need to be included in the analysis.
