Monday, May 5, 2014

Models

Trying to learn different models because my DV in this paper is a panel-count variable.  Doesn't help that I'm an idiot who doesn't know shit.

I'm thinking maybe a poisson model but that has some weird assumptions that I'm not sure I can get around.  Maybe a negative binomial?  Of course all of these would be under xtset in stata.  Yay.  Doing stuff in which I have no fucking idea what I am doing.

God I really wish we had already had the Maximum Likelihood class.

Looks like I'm going to need to go pay a visit to a certain prof.  Hopefully he'll be around tomorrow, and hopefully I'll have some time to talk to him between classes.

Though what is cool is that if I do a model that not many people know, I think the chances of my getting published increases, because journals eat that methodology shit up.


Also fuck cutesy shit on facebook.  Fuck all of that noise.


[Edit] - I THINK I BROKE STATA WHAT THE SHIT IS HAPPENING.  (On iteration 3800 for this model...my data only has 34 variables what is happening I don't know how to do negative binomial regression wahhhh.)

No comments:

Post a Comment