- Joined
- Nov 20, 2011
- Messages
- 54
- Reaction score
- 0
I am using an ordered Probit model (http://en.wikipedia.org/wiki/Probit_model, http://en.wikipedia.org/wiki/Probit) but I may end up using multinomial logit (http://en.wikipedia.org/wiki/Multinomial_logit).
An initial econometric analysis of data actually tells me that
Puerto Rico >>> WV/AR/NE/AL/KY/LA/MS/DE
> NY/PA/TX/OH > CA > all other states.
Of course, all this is very, very tentative! I haven't incorporated the WICHE states appropriately but it already looks like Texas isn't as great as it is made out to be. The above analysis is for the White applicant. Acceptance probabilities for minority applicants will be different.
For a more accurate model and analysis, I first need specific data beginning with:
< State, Total # of MD seats, # MD seats for instate applicants, Total # of DO seats, # DO seats for instate applicants >
Any idea where I can get this data for every state, especially the italicized bolded ones above?
Can anyone email me their MSAR?
An initial econometric analysis of data actually tells me that
Puerto Rico >>> WV/AR/NE/AL/KY/LA/MS/DE
> NY/PA/TX/OH > CA > all other states.
Of course, all this is very, very tentative! I haven't incorporated the WICHE states appropriately but it already looks like Texas isn't as great as it is made out to be. The above analysis is for the White applicant. Acceptance probabilities for minority applicants will be different.
For a more accurate model and analysis, I first need specific data beginning with:
< State, Total # of MD seats, # MD seats for instate applicants, Total # of DO seats, # DO seats for instate applicants >
Any idea where I can get this data for every state, especially the italicized bolded ones above?
Can anyone email me their MSAR?
Last edited: