xerox wrote:
Even then your GPA and school name are not stand-alone deciding factors. Imagine you could attach a numeric value to different parts of your application, and you could run a multiple regression to find out the effect of a hundred different things on the outcome of the adcom's decision. The conclusion will most probably be that both the GPA and the name of the school would each only have very small R-squareds, and therefore will only contribute a little bit if at all.
That analysis is making the classic mistake of analyzing a limited pool of data and drawing conclusions that are too broad. I completely agree that the r-squares are small in both cases, but you're getting lost in the numbers. Yes, there are a multitude of factors (someone could probably identify dozens if not hundreds of relevant factors if they were inclined) and most have very little value, but GPA and school are not among these.
GPA and school though, should not be confused with the muck. The data missing from the analysis are the multitudes of people denied admission. To ascertain the importance of GPA and school, you can't just look at a sample of admitted students, which no doubt have a disproportionately high rate of high GPAs and a very limited sample size of schools. A more relevant sample would be admitted and denied students, as well as those that self-select and eliminate themselves because their GPAs are too low and/or their schools don't carry enough weight, and possibly even those that didn't even attend or graduate from college.
If you factor in the thousands each year who would love to attend a great business school (let's just limit it to college grads), a certain percentage, perhaps 50%, perhaps 80%, don't even consider applying because they don't have the grades or academic background. Take an informal poll of your college friends and think about how many would love to be in a position to apply to a top school. I'd say at least 50%, perhaps a lot more, are totally out of the running because of grades (I went to UCLA, which is a feeder school for most of the top business schools). But if you look at colleges in general, I'd say that 80% of college grads don't even consider top business schools because their academic profile is not competitive. That's a way more important factor that the r-square statistics for the admitted students.
So going back to the original point, GPA and school are both important. If you went to an elite school, then an average GPA might be enough. If you went to an average school you might need an elite GA to be competitive. If you went to a below average school, perhaps only a top 1% GPA will make you competitive. One should not analyze the admitted students pool and be confused that GPA and school don't matter because by and large all have very high GPAs and most come from a short list of top schools. The selection factors (self selection and admit standards) ensure that among admitted students these factors will be small - because they will largely be uniformly high and elite. Obviously though, a very large proportion do not have these factors and for these people, the deficiencies are extremely important.