ABSTRACT
The process of data checking can be tiresome, but ignoring it may lead to drawing the wrong conclusions. The best advice is to think carefully about data before get anywhere near the analysis stage — both in the planning of a study and during data collection. Piloting data collection instrument before the study begins can help to minimise ‘bugs’ and possible misunderstandings by both participants and researchers. This includes checking that the data have been accurately recorded and entered. It is helpful to ‘eyeball’ the data to check that the data set looks right. If data are normally distributed, we can use parametric statistical tests to analyse the data. For data that are not normally distributed, there are various broadly comparable techniques called nonparametric tests — for example, the Wilcoxon signed-rank test is a non-parametric equivalent of the paired t-test.
