Statistics at square zero: a survival guide

In order to understand or conduct research, it is important that all clinicians have at least a basic grasp of the commonly encountered statistical tests. Each has its limitations but there is usually a best or more appropriate test to use for a given set of data. This is a key skill for all clinicians and is one of the curriculum competencies identified by the Royal College of Paediatrics of Child Health. Much to the chagrin of many trainees it is also tested in the RCPCH examinations. A basic understanding of some of the more commonly encountered tests is helpful: it allows a more thorough appreciation of the data that are being presented in published studies and for those undertaking the RCPCH Theory and Science examination offers ‘marks for free’. This short article provides simple guidance on how to pick the best test to use for different types of data. It is aimed at a novice who might consider themselves to be at step zero rather than step one.