<p>So I went through the CC Harvard 2012 thread where applicants listed whether or not they were accepted and counted up the numbers and came up with the following results:</p>
<p>Obviously it tilts in huge favor towards being accepted, but there are a few factors we should take into consideration. 121 is a large representative group, although CCers who got in are more likely to post about it as opposed to those CCers who didnt get in. Either way, this should lift most our confidence because we are the cream of the crop :)</p>
<p>I take it you haven't studied AP statistics. The sample is only "representative" if it has the same characteristics as the overall population being described via the sample. That's most unlikely, in a case such as this one where the sample is not a simple random sample.</p>
<p>We're not very representative of the general population, or even of the typical applicant, I would say, but that group of 121 kids is probably a good indicator of this year's yields :)</p>
<p>I'll remind the high school students reading this thread that "representative" is a word that has a particular meaning in reference to statistics, </p>
<p>and it was natural for readers of this thread to think that meaning may have been intended above. Harvard has a strong statistics department, and it surely admits a lot of young people who have studied statistics before applying to Harvard, so that's the usage I was expecting.</p>
<p>He is saying that a sample can be deemed "representative" only if it accurately reflects the population it purports to represent. If the sample has say a higher acceptance rate, higher average stats, or some other skewed distribution of characteristics compared to the overall population of Harvard applicants, then it is not "representative."</p>
<p>^^No, because as you noted yourself it is tilted heavily in favor of being accepted. Logic suggests that accepted students will be far more eager to rush here and post their stats than will dejected applicants who did not get in. Thus the database you have compiled will very likely suggest that CC applicants have far better chances than they actually do.</p>
<p>Yep, that was my point. Self-reported data almost always has that problem: the people who care to self-report are likely to differ in some systematic (but unknown) way from people who don't self-report.</p>
<p>This makes me feel good because I've been on cc so much that I've kind of forgotten that most of the people here are the exception rather than the rule. <em>crosses fingers</em></p>
<p>If you compare the decisions folder with all the chances and other threads, you'll see a lot of posters don't post their results. My guess is that those are mostly the rejects.</p>
<p>Yeah this is what I figured - the authority with which OP spoke led me to believe that there was a list of 120 or whatever people before admissions and they all reported afterward. Unfortunately, this isn't the case. But to say that 51% of CC could get into H is not the biggest stretch in the world - it's probably high 20s/low 30s, but make no mistake about it I'm sure it's up there.</p>
<p>Yeah sorry i never took stats and realized this later (I had heard about voluntary response bias before). I got a bit excited, but I still think it's way above the average 7%</p>