<p>Wow, I must really be a nerd. Okay, so I got d'frd at Yale and I don't feel like posting my own stats. BUT I was interested in what scores looked like for my top four schools. SOOOOO I added up and averaged the SAT scores for all students who said that they were accepted and decided to share it with everybody. REMEMBER I might have made some mistakes here so go ahead and check my work. </p>
<p>Here were the rules. I looked only at scores for students who were accepted. I ignored ACT scores and used only reported SAT scores. If someone took the old SAT I, I used the verbal and math scores and then took the writing score from their SAT IIs. If they gave only their total score (and not the breakdowns) for the new SAT then I used it to average the totals but left that person out of the individual category averages. There were a couple of entries that were jokes so I left those out as well. I went through as many posts as available through the number shown for each school in the "Official Early Action Results Threads".</p>
<p>Stanford = Analyzed through post # 206
Harvard = Analyzed through post # 406
Princeton = Analyzed through post # 346
Yale = Analyzed through post # 261</p>
<p>Remember that since these are averages they might change if more "accepted" posts are added past the ones I had available. Also, its impossible to know whether some admitted students were more likely than others to post.</p>
<p>Now don't jump all over me for being a nerd or trying to push one school. I'd go to any of them. These are just the numbers and I might have made some mistakes.</p>
<p>AND...[drum roll]...here are the results:</p>
<p>Highest Overall SAT I Average for Admitted Students</p>
<p>There are so many problems with those 'statistics...'. First off, you're using only stats of people on college confidential - that alone is enough bias for the numbers to be worthless. In effect, all you've proved is that "People who frequent CC and got into Yale early have lower SAT scores than People who frequent CC and got into Stanford".</p>
<p>Beyond selection bias, you also are using the mean scores, when a better indicator might be median scores - obviously, this is because a single or a few outliers could easily skew a school's mean.</p>
<p>Also, the numbers are already wrong, primarily because I didn't post my SAT scores. Not everyone on CC even posts their stats. Some simply say "I got in!". Perhaps there's a trend?</p>
<p>Also, I'd be inclined to say that CCers tend to be more competive and thus score differently on SATs than the rest of the population.</p>
<blockquote>
<p>...all you've proved is that "People who frequent CC and got into Yale early have lower SAT scores than People who frequent CC and got into Stanford".<<</p>
</blockquote>
<br>
<p>He hasn't even proved that. Given the error inherent in the test score in the first place, the small sample size confined to non-scientifically selected CC posters, the probable inaccuracy of some of the self-reported scores (i.e. some posters are not completely honest), I think it's pretty clear that the differences reported between the schools listed is insignificant.</p>
<p>In this flawed analysis, 2286 and 2271 are merely noise around the same number, or stated better: 2286 and 2271 are two indistinguishable points contained within the same ill-defined range. The actual order of the schools may well be exactly reversed from what is listed here, or maybe Princeton or Harvard are actually on top. It's impossible to tell from the data available.</p>
<p>Hey you guys, I know that this isn't statistically valid. It was just a way of seeing how my own scores compared to those of the reporting CC'rs who got accepted. Of course it relies on honesty in reporting and there's probably a lot of bias. The bias is probably about the same for all the schools though so a rough comparison is still kinda interesting. Mostly, I guess it says that these numbers are pretty high and the top schools are pretty close together. By the way, I wasn't trying to 'spam' the boards! geez, I just thought it would be fun to share. It's kind of interesting though that some stereotypes show up, like some schools accepting more CC'rs with high math scores.</p>
<blockquote>
<p>Hey you guys, I know that this isn't statistically valid.<<</p>
</blockquote>
<br>
<p>Good, I'm glad you recognize that.</p>
<br>
<p>As a rough comparison it is worthless, and it's interesting only if you find invalid numbers somehow interesting.</p>
<br>
[QUOTE=""]
<blockquote>
<p>It's kind of interesting though that some stereotypes show up, like some schools accepting more CC'rs with high math scores<<</p>
</blockquote>
<br>
<p>You can't say whether those stereotypes show up or not. The numbers you are basing it on are, as you admitted, invalid. That is exactly the sort of conclusion you cannot draw from these numbers.</p>
<p>Another thing it shows it that clearly a perfect SAT is meaningless-(ish).</p>
<p>People post on CC they only got a 2200 or something and are told they better re-take it and improve, but when it comes down to who gets accepted- those 50-100-whatever points don't really seem to matter.</p>
<p>Woah, woah. Calm down on goHYPS. Notice how s/he never said that it was "statistically valid." And never does s/he tell anyone else how to use the info. It's just what goHYPS decided to share.</p>
<p>If I wanted reliable statistics, I'd look in a Harvard/Yale/Princeton/Stanford guidebook. Notice how everyone who responded to this thread had to have actually opened it up, rather than have consulted a HYPS guidebook. This is just an observation. Just take some things at face value, rather than needlessly (and heartlessly) tearing into people.</p>
<p>And lay off the mean-juice. Happy Holidays.</p>
<p>I apologize. I wasn't meaning to be harsh. I just get alarmed it when people misuse statistics. Anyone smart enough to be headed for HYPS or other good schools has got to realize, hopefully sooner rather than later, that you cannot draw valid conclusions from invalid numbers. In fact, you cannot draw ANY conclusions from them.</p>