Hi everyone,
I am currently completing analysis for a sample size of around 5,800. When conducting initial normality tests my data is returning as significant (p < .05), including large deviations from the normal range of skewness and kurtosis, with a number of outliers (which are not able to be deleted from the data set). The study is in relation to engagement in a community-based sport.
I was wondering what non-parametric tests are the most appropriate to use? I am testing differences among four groups (g1, g2, g3, g4) and also as an entire sample.
Any advice would be greatly appreciated...