Dear All,
This is my first post, so please bear with me if this question has already been resolved in one of the posts on this forum. I tried searching for it but could not find answers to my doubt.
I am trying to understand the impact of the for-hire ride-hail services on road safety outcomes (total crash, fatal and severe, pedestrian, drink_driving) using two years of data a) the year before the services started and b) the year after it started. I have the annual count of road crashes (segregated by crash types), the location of each crash, the road network, and traffic volumes on each road segment. I also have the bifurcation of traffic volumes, i.e., whether it is a ride-hail trip or not, and the total number of trips starting or ending on each road segment. Given the traffic volumes are annualized, I am using annual comparison as my unit of analysis at a segment level. So, comparing the road crashes in yr1 to road crashes in y2 for the same road_segment_x.
Since the LHS is a count data (crashes), I am using the xtpoisson FE model.
My question is, what is the difference between the exposure variable and the offset variable? Do they take part in the equation model and further analysis? Would they participate in the model equation like the one in excel?
This is my first post, so please bear with me if this question has already been resolved in one of the posts on this forum. I tried searching for it but could not find answers to my doubt.
I am trying to understand the impact of the for-hire ride-hail services on road safety outcomes (total crash, fatal and severe, pedestrian, drink_driving) using two years of data a) the year before the services started and b) the year after it started. I have the annual count of road crashes (segregated by crash types), the location of each crash, the road network, and traffic volumes on each road segment. I also have the bifurcation of traffic volumes, i.e., whether it is a ride-hail trip or not, and the total number of trips starting or ending on each road segment. Given the traffic volumes are annualized, I am using annual comparison as my unit of analysis at a segment level. So, comparing the road crashes in yr1 to road crashes in y2 for the same road_segment_x.
Since the LHS is a count data (crashes), I am using the xtpoisson FE model.
My question is, what is the difference between the exposure variable and the offset variable? Do they take part in the equation model and further analysis? Would they participate in the model equation like the one in excel?
Comment