ABSTRACT
The statistical techniques that we have examined thus far are all parametric statistical techniques. There are two situations in which non-parametric statistical techniques are required. The first situation, as mentioned in Chapter 4, is when the data are not normally distributed. The second set of circumstances, which often occurs in combination with the first situation, is when the data are measured on a scale that does not involve equal units. This is contrary to most unit scales that we are used to dealing with. For example, the interval between 1 kg and 2 kg is the same as the interval between 99 kg and 100 kg. This assumption cannot be made for other scales used within sport and exercise science, particularly those based upon the subjective ratings of subjects. The ratings of perceived exertion (RPE) scale is a good example of just such a scale. It is difficult to know whether the interval between 6 and 7 on the scale is the same as the interval between 19 and 20. It is also open to debate whether data collected using the Likert scales to measure emotional state during sport should be analysed using parametric statistical techniques unless they have been validated as interval scales. Previous research and experience have shown that such scales do often act as interval data, but the question should always be asked whether or not parametric statistics are appropriate for your data.
