On most Ethics Committee application forms there is a question asking how the size of the study was determined. This is an important question as study size has ethical implications. Studies that are too small tend to give inconclusive results; they waste money and time, and study participants are inconvenienced. They can give inaccurate answers, which jeopardise future research. An apparently large effect found in a small trial may make a further trial unethical. Conversely, important true effects may be missed.

Usually a formal study size calculation has to be made and the answers to this question on the form are amongst the most complicated to understand. Sometimes confidence intervals are used to explain the study size (usually for descriptive rather than comparative studies) and these are discussed at the end of this chapter. Formal study size calculations are used when the purpose of the study is to compare two or more groups of subjects. There are five areas that must be considered:

• Difference of interest

• Variability of measurements

• Significance level (p-value)

• Power of the study to detect a difference

• The proportion of the study population that is ‘exposed’ or taking the treatment

People without a statistical training may find it difficult to assess the study size statements and will have to rely on a statistician. However, they can assess whether the difference of interest seems reasonable.