I have a nominal data set of a sample (species) floating characteristic (vertical floating, horizontal floating, bottom horizontal, and etc.) over a period of time.
I'm interested to study the most dominant floating characteristic and the pattern/changes of floating orientation within a period of time.
Other than observing the mode of the most observable and longest floating characteristic, is it suitable to calculate the error of binomial/multinomial proportion? Do suggest any test that might be useful for my study as well.