I have multiple behavioural categories (Foraging, vigilance, walking... etc) for mean percentages of each behaviour performed for each observation period (12 total).
Sahar Firoozkoohi ANOVA can only be used on normally distributed data - which mine is not.
I also do not have multiple habitats. I have been observing geese at the same location over 12 observation periods, and have recorded their behaviour at fixed intervals of 5 min.
Sahar Firoozkoohi The tricky thing is that I do not meet the assumptions of Kruskal-Wallis because my behaviours are not truly independent of each other. A goose performing one behaviour cannot perform another behaviour at the same time.