Myths and reality about calculating sample size

When we decide to conduct a study, one of the first questions that arises is what number of individuals should be included in the sample for it to be `representative' and for the study to be `valid'? As in other areas of life, there are many matters for which there is no `right' amount and different quantities are valid. The same applies here. When asked the question `How many euros did this bicycle cost?', the answer is a definite number. But the question `How many euros do I need to buy a bicycle?' can be answered in many different ways, depending on the size and other characteristics of the bicycle. Statistics textbooks contain formulas relating sample size to certain parameters and most doctors believe that one of these will give them the `right' size for their research, and that by using them their choice of sample size will be justified in the eyes of potential reviewers. This document reflects on the true value of these formulas and how researchers should make proper use of them. It is necessary to show errors and simulations that benefit no one and hinder many by taking up large amounts of time and energy.