<p>
</p>
<p>I don’t see a problem with the size of the sampling pool. AP/Gallup polls are often conducted to get a view of the opinions of the entire United States of America with a sample of 1000 people. Here, we’re looking at something like 100 people out of 3000, which isn’t too outrageous.</p>
<p>
</p>
<p>Not necessarily. First of all, I’m not all for the theory that CCers are so different from the normal applicant. We’ve seen plenty of deferred and rejected applicants, and probably only a slightly higher percentage of accepted applicants than would be if we selected a random group of applicants.</p>
<p>Secondly, even if we assume that CCers are all outliers, which is really pretty ridiculous, you can still look at patterns WITHIN this group to see admissions trends. This was the point of the above link.</p>
<p>
</p>
<p>Look at my previous example of AP/Gallup polls. Those kind of polls are almost always correct within 5 percentage points, even though you’re looking at less than .001% of the population.</p>
<hr>
<p>Perhaps disconnectedly and unprofessionally, I’d like to say that I think the results here are pretty obvious. One can always come up with a somewhat rational argument to refute any other argument, but I don’t think you can seriously say that there was no drastic change in admission from last year to this year when you compare the EA threads of 2011 and 2012. I’ve seen people argue against the apparent fact that affirmative action is used at leading institutions like MIT. After all, there’s really very little non-anecdotal evidence of affirmative action taking place, is there? MIT doesn’t publish statistics like that, and they never will. So how can you judge on whether there is AA at MIT? Anecdotal evidence, personal experience, and common sense.</p>