Cheers; I am examining student drop out rates from year 2000 to year 2010. I read that Poisson regression is appropriate method for such an approach. Ultimately the goal here is to compare if there are differences in drop out rates between students who entered in year 2000, year 2001, all the way through students who entered in year 2010. What I am confused by is how to structure my data. Here is what I have currently just as an example
student entry year drop gender gpa
1 2000 2000 0 1 2.00
1 2000 2001 0 1 4.00
1 2000 2002 1 1 0.00
1 2000 2003 1 4.00
2 2000 2000 1 1 0.00
2 2000 2001 1 1.00
2 2000 2002 1 1.00
2 2000 2003 1 3.00
3 2000 2000 0 2 2.00
3 2000 2001 0 2 1.00
3 2000 2002 0 2 3.00
3 2000 2003 0 2 1.00
4 2001 2001 0 2 3.00
4 2001 2002 2 2.00
4 2001 2003 1 2 0.00
4 2001 2004 0 2 2.00
5 2001 2001 0 1 4.00
5 2001 2002 0 1 3.00
5 2001 2003 0 1 1.00
5 2001 2004 0 1 4.00

Code:
 
 poisson drop i.entry gender gpa, vce(robust)    
 poisson, irr
Is this correct? Again I just want to compare drop out rates between students who entered the school at different years and then see how gender and gpa impacts this. Or do I need to calculate rates and then run this model?