hi everyone, I want to use annual road traffic accident data for analysis. Since the people who cause car accidents are different each year, it is not so-called panel data. I currently think that use logistic regression or negative binomial regression for analysis maybe can work.
I want to focus on a policy effect which passing a car driving license test needing add a road driving test. I want to compare the impact on traffic accidents before and after the law revision.
the following is my thought that i want analyze:
1.The number of deaths/injuries/total deaths and injuries caused per road traffic accidents
2.No serious injuries, minor injuries, serious injuries, deaths
here is my question:
1. If i use logistic regression or negative binomial regression, can i use the did method for analysis to analyze the control and experiment group the difference before and after the law revision
2. In addition to the above two methods, is there any other feasible method?