Using randomization techniques to analyse behavioural data

摘要: Abstract Data from behavioural studies are frequently non-normally distributed and cannot be analysed with traditional parametric statistics. Instead, behaviourists must rely on rank-transformation tests, which lose potentially valuable information present in the data. Recently, however, biologists other disciplines have resolved similar statistical difficulties by using resampling methods. Results Kruskal–Wallis non-parametric ANOVA randomization tests were compared for two data sets. It was found that more powerful than Kruskal–Wallis, could thus detect smaller effect sizes In addition, variance calculated around theP-value at eight levels of replication ranging 500 to 10 000, determine optimal number replications test. The decreased as increased. TheP-value stabilized 5000 replications, it is recommended least used