If dependent variable - the number of participation in legislative elections (1-8), which regression is better? count model and negative binomial regression? If 'yes', why?

Similar questions and discussions