Comments on: Earmarks and the house that approves them… http://thespicychicken.com/2009/08/23/earmarks-and-the-house-that-approves-them/ n'th generation research Mon, 14 Jun 2010 21:23:20 +0000 http://wordpress.org/?v=2.9.2 hourly 1 By: Nick http://thespicychicken.com/2009/08/23/earmarks-and-the-house-that-approves-them/comment-page-1/#comment-323 Nick Sat, 29 Aug 2009 21:49:20 +0000 http://thespicychicken.com/?p=143#comment-323 Actually, you don't have to. When you are testing if the means are different from two populations all you are really looking at are the mean and standard deviation (in a t-test at least). So the sizes don't have to be equal to test this. The problem is one of confounders. I don't have an estimate for how much easier it may be to get earmarks passed for dems vs reps due to the dems controlling the house. So there may be a bias that we would need to estimate, which we could, if we had this data from when the reps controlled the house. Actually, you don’t have to. When you are testing if the means are different from two populations all you are really looking at are the mean and standard deviation (in a t-test at least). So the sizes don’t have to be equal to test this.

The problem is one of confounders. I don’t have an estimate for how much easier it may be to get earmarks passed for dems vs reps due to the dems controlling the house. So there may be a bias that we would need to estimate, which we could, if we had this data from when the reps controlled the house.

]]> By: Jason http://thespicychicken.com/2009/08/23/earmarks-and-the-house-that-approves-them/comment-page-1/#comment-322 Jason Sat, 29 Aug 2009 21:25:12 +0000 http://thespicychicken.com/?p=143#comment-322 Interesting! Can't you do some sort of controlling for the # of dems vs #of reps ? Interesting! Can’t you do some sort of controlling for the # of dems vs #of reps ?

]]>