Outliers (diagnosed with Grubber’s Test) were handled in two
different ways (Cousineau and Chartier, 2010): (1) they were replaced
by the mean of the group in the repeated measures ANOVA
to avoid losing a whole case; (2) they were removed from the analysis
in the non-parametric analysis and in the one-way ANOVAs.
Initial sample sizes were as follows: sham-operated males = 10;
urinary bladder damaged males = 9, castrated males = 9. However,
sample size may vary across days because of the removal of outliers.
One sample was lost during T analysis.
The effect of treatment on courtship and male aggression towards
females was analyzed by using the mean value of post