I wish I could do a Trackback on blogger (if anyone knows how please leave a comment). I want to respond to one of Brit-Chicks posts.
Let me start by saying that this post is quite good, and that overall I agree with much of what she is saying. However I think there is a subtle point about the statistics that she didn't get quite right. :)
You can read the article
here, it's the post with the heading "Lies, damn lies and statistics".
The point I want to address is the concept that if total applications are lower then it's easier to get into business school. Her thought was that unless the total number of applications drops below the number of open positions for the school, she didn't see how it would be any easier to get into school. I want to show that a simple decrease in applications can increase your chances of getting in.
Ok, let's consider a typical application process. Wharton is good enough to post this information for us, so I'll use their numbers. You can find them
here. In 2004 there were 7200 total applicants to Wharton of which they admitted 1,130. This number works about to be 15.694% of the people who applied were admitted. However of the 7200 people that applied roughly 20% to 30% of them were not academically qualified to attend Wharton (According to FF and Alex Brown). Let's take the 20% number just for the sake of argument. That means there were really only 5,760 qualified applicants to Wharton, of which 1,130 got accepted or 19.618%.
Wharton is shooting for a class size of around 800, by managing their wait list properly and by understanding yield, they successfully crafted a class of 799 students.
Now let me present the crux of my argument. Let's say for the sake of this discussion that Wharton ranked all graduates from top to bottom with no ties (they don't do this from what I understand). So we now have a ranked list from 1 to 7200, anyone with a number above 5760 wasn't academically qualified to attend Wharton. However the people with numbers 1 through 1,130 were the best qualified (again for the sake of argument only) and thus were offered a spot in the class. Let's further assume that Wharton is aiming for a fixed class size of 800, and they know that traditionally about 29% of the people offered a position will go to a different school. To arrive at the number of offers the school should place we use the following equation (100% - 29.20%) = 70.8%, plug the percentage in as follows: .708x = 800 and solve for x. x is approximately 1130 in this case. Now let's assume that in this particular year you were ranked 1131, and just missed the cut.
Now to illustrate my point, next year lets say only 6,000 people apply to Wharton instead of 7200. Assuming 20% of the applicants were again not academically qualified, we have a list of 4800 applicants that are qualified to attend Wharton. Again for the sake of argument, Wharton is using the same ranking methodology, and again they expect the same yield so they know to issue 1,130 offfers to craft a class of 800 people. Based on this we can see that if 1,130 people out of 4,800 are accepted, the acceptance rate is 23.541% versus 19.618% for the previous year. We can see that the odds of getting accepted are now higher than the previous year.
The crux of my argument and perhaps it's weakest point, is that people are ranked for admission. From what I understand the process of admission is much more complicated than this. However if you assume that there are more qualified applicants than there are slots, the admissions committee uses *something* to make a determination about who will get accepted and who gets rejected. This means that somewhere along the line they are saying I like candidate X better than candidate Y, even though both of them would fit at Wharton. If however candidate X doesn't apply, then canditate Y gets admitted instead.
Another problem with my argument could be the distribution of the candidates that don't apply. For instance if for some reason all 1,200 of the people that didn't apply were ranked below you on the list, it wouldn't help you to get accepted. I don't know what the breakdown is, but I would expect that at least some of the better applicants won't apply, therefore increasing your odds, if not exactly by the numbers shown above.
My point isn't that decreased applicants necessarily increase your chance, but that it can increase your chances depending on the situation. This post is in no way meant to slam on Brit-Chick. I enjoy her blog immensely, I just wanted to clarify this one point. :)