Fancy Pants wrote:Stringer Bell wrote:I agree with your assumptions, and this list obviously has a high probability of looking different with '09 clerkship data, but I think assuming static clerkship placement is more useful than just looking at '08. I work on an analytics desk and we make these kinds of assumptions internally in the absence of actual data when we have to. You just have to understand you are dealing with incomplete data.
The conclusion that you can draw from this is that the '09 placement breakout by school MIGHT look different than most of us thought it would.
So in your job they think it's a good idea to assume that data will remain consistent from one year to the next even when there are major reasons for believing things will be very different, and when the release of similar data has already confirmed that things will be very different?
I'm no analyst but that seems like a stupid idea.
Also, just FYI. Wrong data =/= "incomplete data".
In all honesty we would try to dig further and find out the answers to questions like what % of each school that tries for clerkships gets them and extrapolate that across the theoretical % of no offered students that will be looking for clerkships. For example, I might take a clerkship or two from each school and give it to Harvard. My point was that to predict what '09 looked like, I would rather have data for one variable and make a very crude assumption for another piece than just look at '08.