<p>
</p>
<p>Jersey13 is talking about admitted students. None of this discussion of percentiles is about students who are applying. It’s about students who are admitted. Can we agree that a significant percentage of students at Ivies are hooked? Can we agree that a significant percentage of students at Ivies is not students with high SATs and nothing else? There we go, unhooked students should be looking at higher than the average.</p>
<p>Commence statistical junk: Jersey13’s statistics are also better than sic’s. S/he suggests that we use the 75th percentile because only that and the 50th percentile are published. You, sic_infit, suggest that we average the 50th and 75th. That’s not how bell curves work, assuming that Ivy admissions have at least an element of the bell curve to them. (Probably chopped off after the first or second standard deviation on the high side, due to there being a maximum score on the SATs.) Because a bell curve becomes less dense farther from the mean (the 50th percentile), moving halfway, scorewise, between 50th and 75th means that you’ve captured a much larger population below your median score (and above the 50th) than above it (and below the 75th). This would be easier if I could draw you pictures…Regardless, the statistical implications get wonky really quickly, so just going with “I’m aiming for the 75th percentile or a bit below” is ideal. Not actually trying to get an estimate with false accuracy of where precisely that bit below falls.</p>
<p>Sorry about the statistics rant, I guess a little bit of that class did rub off on me… (despite my best efforts to the contrary lol)</p>