So I have a presence/absence (1/0) dataset and right now I am trying to find out which model to use. My data looks like this:
Dependent var: presence/absence (95% zero's)
Independent var: month (1:12), hour (0-23), location and individual (random in mixed model).
Since the data consists for approx. 95% of zeros, I started to wonder if there is something like 'zero-inflation' in logistic regression?
Best Answer
No. There are only two outcomes in logistic regression, adding a zero-inflated component is pointless. Also, having many zeroes does not imply zero-inflation. Zero-inflation means more zeroes than expected under your model. So first fit the model and you'll see that it will manage to deal with the zeroes.