<p>For the record, I'm only posting this because I find the information interesting. I am not arguing against anybody. I am also not being arrogant. Cheers.</p>
<p>Incidentally, MIT's suicide rate is not low, but it is well toward the middle of the pack - it's definitely not bad.</p>
<p>From the excellent link which was so kindly presented to me:</p>
<p>"Examine MIT admissions in Figure 1. The probability of a student's being admitted rises steeply and monotonically in her or her combined SAT score, suggesting that MIT is not engaging in strategic admissions. Now examine Harvard admissions in Figure 1. The line has a flat region that suggests that the probability of a student's being admitted is about 10 percent regardless of where his SAT scores in the range between the 93rd and 98th percentiles. Above the 98th percentile, a student's probability of admissions rises steeply. Finally, consider Princeton admission in Figure 1. At Princeton, the admissions probability rises to 20 percent at the 93 percentile, then falls to 10 percent at the 98 percentile (precisely the region where competition is toughest), and then rises again for students with SAT scores in the top 2 percentiles. In short, it appears that Princeton practices more strategic admissions than MIT or Harvard.... Such behavior is potentially costly to the actual quality of an admissions class, with no clear benefit beyond a higher reported matriculation rate.... Princeton alters its acceptance decisions in order to avoid match-ups with Harvard, Yale, Stanford and so on. We would be unable to rank Princeton rank vis-a-vis its close competitors because its match-ups would always be against less selective colleges."</p>
<p>MIT also has a more self-selective application pool. Read the part about how the admission officer at Duke specifically talks about the mandate being to increase their rankings by advertising less stringent criteria than it actually applies, and encouraging early decisions applicants.</p>
<p>Additionally, it is known that MIT wins more head-to-head admissions battles against Harvard than any other school, which severely contradicts the results in the paper, which are given only as "an example of a ranking" as were based on a monte-carlo simulation from only 4000 data points. Also, Caltech has a more specific admissions pool, which pretty much only overlaps with MIT. MIT's pool overlaps more with HYSP.</p>
<p>If I had a nickel for every person who asked me about the suicide rate at MIT, I would be a very wealthy woman.</p>
<p>Has MIT had people commit suicide? Yes. Could the mental health facilities be better? Probably. Is MIT really all that different from any other school? No.</p>
<p>People like to trot out stats like MIT's suicide rate is 10.2/100,000. The problem is that on a campus with about 10,000 students (grad + undergrad), that works out to just over one suicide per year. (Which matches my experiences there.) If MIT were to get two in one year, the suicide rate for that year would skyrocket to around 20. In other words, with such a small sample size, you can get wildly varying numbers.</p>
<p>The truth is that MIT's suicide rate isn't outside the norm for Americans between the ages of 18 and 25. It's just that people have this preconceived notion that such a tough place will drive people crazy and then that prejudice is reaffirmed when they hear about a death on campus. </p>
<p>Basically, if other colleges were as news worthy as MIT, you would hear about more deaths on those campuses as well.</p>
<p>Rooster said "Just to let you know, however, more people on average choose Caltech over MIT when admitted to both."</p>
<p>Don't flame me; I think Caltech is an awesome place, and this isn't meant to put it down in any way. I feel compelled, however, to correct the above statement because it is utterly false. I haven't read enough of the pdf to determine how this "Revealed Preference Ranking" was calculated; what I can give you instead is a simple summary of our cancellation study.</p>
<p>Of our ~1500 admits this year, 1390 responded to our cancellation study (92%). The below numbers aren't exact because we didn't get 100% to respond, but 92% responding is enough for me to draw some conclusions.</p>
<p>Of the 1390: 178 people were admitted to both Caltech and MIT. 64% of them chose MIT, 19% of them chose Caltech, and 17% chose neither.</p>
<p>You can massage the data to yield your "Revealed Preference Ranking" however you'd like, but the raw data shows that cross admits choose MIT over Caltech 3 to 1.</p>
<p>I just used the fact that the Revealed Pref ranking put Caltech higher than MIT. This may mean that Caltech is more successful than MIT overall in winning cross-admits from many different top schools like Harvard, Stanford, Yale, etc. </p>
<p>However, isn't Caltech's student body a heck of a lot smaller than MIT's? I don't know the exact numbers, but I think the number of Caltech-MIT cross-admits may in fact exceed the number of spaces in Caltech's freshman class. This is not the case with MIT.</p>
<p>Given class-size considerations, 19% of cross-admits may be a very significant percentage of Caltech's incoming class whereas the 64% that MIT wins may not be that significant at all.</p>
<p>MIT's suicide rate is not necessarily reflective of how mentally unstable the students are. Keep in mind that MIT's curriculum is insanely difficult, with no grade inflation to help anyone either. It's understandable for those students to be under a lot of pressure. It's not because they have psychological disorders, but rather the fact that they're living in a high-stress environment, much more so than students at other top schools which are notorious for grade inflation.</p>
<p>Rooster, I don't think you can say anything about the relative intelligence of the students based on the 19% that chose CalTech over MIT. Maybe they wanted to stay close to home. (Something like one third of CalTech students are from in state). Maybe they don't like the cold. Maybe they just saw Real Genius and they want to build a really cool laser. =)</p>
<p>
[quote]
Given class-size considerations, 19% of cross-admits may be a very significant percentage of Caltech's incoming class whereas the 64% that MIT wins may not be that significant at all.
[/quote]
We'll have to use the 2004-2005 Common Data Set for class enrollment figures, and Ben's statistics above from the MIT admissions survey for 2005 applicants, which don't quite match up, but it's the best we can do just now.</p>
<p>64% of 178 cross-admits (114 people) chose MIT. MIT's previous year's enrolled class size was 1077, meaning 10.6% of the MIT incoming class, roughly, would have selected MIT over Caltech.</p>
<p>19% of 178 cross-admits (34 people) chose Caltech. Caltech's previous year's enrolled class size was 207 (!), meaning 16.4% of the Caltech incoming class, roughly, would have selected Caltech over MIT.</p>
<p>The number of cross-admits does approach the size of last year's Caltech incoming class. In addition to the factors mentioned by vtardif, some of those choosing Caltech probably did so because of the small size of the school.</p>
<p>Mollie, don't worry. Ironically, I was using the term "psychological problems" thinking it was the "nice" way to refer to the depression and suicide.</p>
<p>And you're right, it can't be evaluated out of context, but in context, it's still a bit frightening. Does that mean that MIT is a bad school that doesn't take care of its students? Of course not. But it does happen, and that's all I was trying to say. It just seemed to make sense to me that when a significant percentage of your student body was #1, a lot of people will have a hard time adjusting to being at the bottom (even though MIT doesn't rank, I know you get reports with your GPA as compared to the average, etc).</p>
<p>That being said, MIT is still my number one choice for my undergraduate pursuits. Things like suicide rate are things you control. I know I won't kill myself.</p>
<p>For the record, the revealed preference paper is dismissible, as they themselves say. They tested randomly 4000 high school students. Of these, it's likely that less than 20 applied and were accepted to MIT, and even fewer to Caltech. Hardly statistically significant.</p>
<p>Also, 64/19 is the meaningful statistic. 10.6/16.4 is meaningless, although it explains why people at Caltech have the feeling that most people would pick Caltech over MIT.</p>
<p>Ben Jones correctly quotes the straight cross-admit numbers. (I myself was going to correct rooster's misinterpretation of the paper, though I'm sure it was an honest mistake.) I am a little embarassed to admit, since one of my majors is econ, that I haven't read all of the NBER revealed-preference paper. I do know that they started with the assumption that there is a mean ranking among academically elite students of academically elite colleges, and then asked the standard Bayesian question "given the actual choices of the students, what is the most likely distribution of the mean desirabilities?" The calculations are very standard and are used in revealed-preference comparisons in industry all the time. Their answer -- based on a data set of N = 3,240 high-achieving students -- is in the paper.</p>
<p>While it's certainly possible that their data was just nutty, it's important to note that straight head-to-head numbers can also be misleading. For instance, it is well-known that even in the stratosphere of elite colleges, geographic proximity plays a very strong role in deciding where students go. I recall hearing that at least 40% of the variation in college admissions decisions among academically comparable colleges can be explained by geographic proximity. So to tease out the actual "prestige" or "quality" factor from the others one has to run regressions to exclude the confounding variables. molliebat and I seem to be striking a common theme, on different subjects -- straight averages don't tell you the whole story.</p>
<p>I would be very curious if the revealed preference ranking were done with a much larger data set.</p>
<p>It is a matter of choice and opinion whether one wants to heed the 64/19 statistic or the 10.6/16.4 statistic. In absolute raw numbers, MIT wins more cross-admits than Caltech. However, as a percentage of the freshman class, Caltech has more people who rejected MIT than MIT has people who rejected Caltech. Both are pretty significant. </p>
<p>It isn't right to penalize Caltech for having a smaller class size.</p>
<p>For the record, I didn't think the 10.6/16.4 statistic was particularly meaningful, I was posting it to answer the speculation that the percentages might be wildly different and highly significant.</p>
<p>Rooster, the second one sounds like it's meaningful, but mathematically it is useless. It's just the first statistic weighted for class size, which is an irrelevant factor here. When people choose from school or the other, that's a boolean, and we're just comparing boolean counts. All that really matters is, out of all people who were admitted to Caltech and MIT, which did they pick. The only added information we get from the second statistic is that as a percentage of class size, Caltech has more people who were admitted to both schools. All this tells us is that Caltech admittees are more likely to have applied to MIT, than are MIT admittees likely to have applied to Caltech.</p>
<p>
[quote]
It isn't right to penalize Caltech for having a smaller class size.
[/quote]
</p>
<p>Wait, I'm not penalizing Caltech for anything - I think Caltech is awesome, as you'll see in many of my other posts. I was just correcting that one statement; I didn't (and don't) have an agenda, truly.</p>
<p>In different threads, Ben G. and I have talked at length about the differences between the schools, which lead to them being "ideal" matches for slightly different kinds of students. I'd guess that this affects the decisions of cross-admits even moreso than location, which is certainly also a major player.</p>
<p>As each school has a slightly different vision of the "ideal applicant," I think the real comparison here is Caltech's ability to recruit and enroll its "ideal student" versus MIT's ability to recruit and enroll its "ideal student." Not that one can easily measure that sort of thing (too many intangibles), but if one could, my guess is that we'd find their success to be virtually identical.</p>
<p>In other words, Caltech is good at recuiting Caltech students, and MIT is good at recruiting MIT students. :) And the world is all the better for having both.</p>
<p>Ben J., you are the voice of reason as always. Especially about how hard it is to find meaningful metrics telling us how well schools achieve meaningful goals. : )</p>
<p>Oh, so true about the evasive metrics! But on the flip side, that huge grey area does keep our jobs exciting and thought-provoking - at least we're not bored! :)</p>
<p>:-P Sagar. I should point out that I'm just a Caltech junior who is on the admissions committee at Caltech, whereas Ben J. is a professional admissions officer at MIT. He is certainly more knowledgeable about "admissions matters" than I am, but it's always a pleasure to discuss things with him here at CC.</p>
<p>I do really appreciate the folks at MIT making themselves available and making the process transparent; some of my colleagues at Caltech and I try to do the same at our school. Hopefully it takes some of the craziness out of this process... I sure wish it had been the norm when I was applying to college a few years back :)</p>
<p>LOL Ben, I respect your modesty but one glance at your body of work on this site alone shows that you could hold your own against any "professional" admissions officer. :) If you need a break between undergrad and grad you should seriously think about doing this for a couple of years - you're already a real asset to Caltech and the admissions world in general.</p>
<p>Sagar - we're just good at creating these images of ourselves. In truth, MIT and Caltech share a large computer (it's physically located somewhere in the middle of the country) that makes all of the decisions. We actually license it to the Ivies to fund our annual trips to Hawaii and Ibiza. :o</p>