<p>Oh ok I didn't realize that</p>
<p>xiggi,
Middlebury dropped the SAT requirement nearly 20 years ago--that doesn't explain the increase this year...</p>
<p>Few thoughts on Holy Cross- the school is nearing completion of capital campaign which is well over original goal and plans to offer more merit aid in the future. Perhaps, HC benefits from a Division 1 sports program as Colgate, Bucknell, and Davidson(basketball only) are some of the few LAC's offering D1 sports. Also Duke and Notre Dame should see application increases next year due to great visibilty via basketball and football respectively.</p>
<p>what is the difference between SAT test takers this year and last year?</p>
<p>does anyone have a count on the number of HS seniors this year as compared to last year? and next year?</p>
<p>I am completely confused. I don't know how an institution can report scores they do not have. Can you walk me through how you'd handle a concrete example of fictional students? A manageable number? </p>
<p>Here are 20 students. 12 submitted scores (60%). I've listed them in order of their scores, with the non-submitters following. </p>
<p>Kurt 1150
Nora 1240
Paloma 1240
Inez 1250
Ellen 1260
Brianna 1300
DeWayne 1340
Juan 1350
Steven 1380
Fiona 1450
Leon 1460
Octavia 1500
Antoine Didnt submit
Cathy Didnt submit
Garrett Didnt submit
Harold Didnt submit
Miriam Didnt submit
Quentin Didnt submit
Rachel Didnt submit
Tiffany Didnt submit</p>
<p>The mean SAT is 1327. The 25-75 span is 1248-1398. This is calculated only on those 12 students who submitted (hence the odd numbers--in essence I'm reporting a point between the 3rd and 4th student, and the point between the 9th and 10th student. With bigger numbers we could land on the exact student, but that would be cumbersome in this example).</p>
<p>Now, am I understanding you right in that you want a quartile that includes Antoine, Cathy, and the others who didn't submit? That is, all 20? That would put the 25th percentile right between the 5th and 6th student, and right between the 15th and 16th student. To do that, as you noted, I have to put all 20 "in order." How do I order them? Do the students without scores get put on the bottom? On the top? Mixed in at random?</p>
<p>Am I misunderstanding you?</p>
<p>I have no idea why this is so hard. If there are 20 students in the entering class, the 75th percentile is something like the 5th student down from the top, here 1340 or 1350. The 10th student down is about 1240 or so. We don't have a number for the 15th student down, and we so report. Statisticians must have answers as to how to deal with the artificially small population here. What's the problem?</p>
<p>
[quote]
hm for a great school like Wm & Mary I would have expected greater increase- have they had to increase tuition lately?
[/quote]
</p>
<p>W&M applications have been increasing at a pretty good clip the last few years. You also have to remember that although a state school, W&M is small compared to the other top public schools, only 1400 students per freshmen class (slightly over 3000 accepted). Tuition has been increasing slightly each year, I believe, though don't quote me on it... I'm on the VA prepaid tuition plan, so it doesn't affect me. Cost is still slightly cheaper (few hundred dollars) than UVA.</p>
<p>The Charter Program is in the works though... and I imagine once it goes through legislature that tuition could see a small but noticeable jump (in line with UVA and VTech raising tuition as well).</p>
<p>eulen, you're assuming the "non-reports" all have scores at the bottom of the pack (or would, had they taken the tests). While you can argue that this may be a sensible interpretation, it has no basis in fact, and provides a list that doesn't mean anything. It seems much more meaningful to provide the 25-50-75 %-ile for those who report scores and describe what percentage of the enrolled class that represents (since a missing score tells you nothing about the individual's place in the ranking). Leave out the non-reports and rank the rest, then do the stats. I think that's what everyone else is saying.</p>
<p>My point is not the artificially small population here. The only reason I made it small so we could actually do the math, the reordering, etc. We agree--in post #65, see the last full paragraph. There I stated that the 75th percentile would be about the 5th person from the highest. That's what you said, too. This is not in dispute. (yes, technically, the quartile falls between two students, but this is immaterial--I addressed that earlier). </p>
<p>My confusion remains, though. Percentiles are based on placing people in order based on the value of the measure of interest. Eight of those students have no value. How can you order them in with the other 20? Are you saying that institutions must assume those eight, had they taken the SAT (or submitted the scores) would all have scores lower than the lowest known value?</p>
<p>"you're assuming the "non-reports" all have scores at the bottom of the pack (or would, had they taken the tests). While you can argue that this may be a sensible interpretation, it has no basis in fact, and provides a list that doesn't mean anything."</p>
<p>OK, you convinced me. The non-reporters were all in the top 10 percentile.</p>
<p>Huh?</p>
<p>I wasn't trying to "convince" you that the non-reporters are all in the top decile. I was stating that you have no facts on which to base an assumption.</p>
<p>As hoedown says, you cannot order the non-reporters in with the reporters. (hoedown, our posts crossed in the howling ether. ;) )</p>
<p>Right. They just wanted to keep their good scores a little secret.</p>
<p>This mom just forked out the money for Physics C, Mechanics, and Calc AB. Son will be taking the tests at a high school in Israel, since the entire senior class migrates there for several weeks in May. Since EA son still isn't sure where he's going, he'll give it one more try.</p>
<p>eulenspiegel - there are students out there who get a 1560 at first sitting and then re-take once or twice to "improve" their score. Should such a student choose to apply to an SAT-optional school, who knows whether they want to submit their scores or not?</p>
<p>I, too, question your assumption that non-submitters would rank, without exception, at the bottom. Schools such as Bowdoin, Bates, Colby attract a very strong student body and admit a very strong student body. Some of these kids may choose not to submit just because it saves a few bucks and one more clerical job.</p>
<p>SAT scores are for me- not a huge source of information about a school.
I also want to know graduation rate- student : prof ratio, size of classes, a feel for numbers of students going on to grad school and students transfering, I want to see what the dorms look like and talk to the RAs. What the labs and the libraries look like, what percentage of students are Greek, what percentage receive aid and how much, what part of the country they are from.....
Really, while SAT scores tell me a * little* it is only one measurement, of one day & while the numbers might tell us if my D is in the ballpark of other students, it isn't going to tell us if she should apply or even visit the school. My daughters GPA and SAT I scores were lower than the average of the school where she will be graduating this spring, so apparently the college felt that it was only * one * piece of information about her as well- thank goodness</p>
<p>The increase in applications to Middlebury probably can be explained, at least in part, by the recent change in required standardized test score submissions. Before this year, applicants were required to submit either an ACT score or scores from exams in 3 different study areas, one of which had to be quantitative (such SAT II Chem, Physics or Math, or a math AP). The math portion of the SAT I exam did not satisfy the quantitative testing requirement. I assume that the math testing hurdle might have discouraged some humanities/social sciences kids (who did not take math/science SAT IIs or AP exams) from applying. Middlebury's application requirements now seem to more closely parallel its peer institutions.</p>
<p>emeraldkity4, your last sentence intrigues me as much as anything I've ever read on this board. After all, each of us is trying to figure out how decisions are made. Would you share with us the factors that you think influenced your daughter's admission to her current school? Were they national awards and/or athletic rankings? Or something less flashy that might give the rest of us more hope?</p>
<p>Eulen, I don't understand the rancor in your posts. No one has suggested they were in the top 10%. If that were the case, that would make this quite easy, wouldn't it? We'd know exactly where they fall, above the 75th percentile. The point that several of us are making is that we don't know where they fall.</p>
<p>I also am surprised at your assertion that these students are "hiding" their SATs (presumably because they reflect so poorly on the candidate). First of all, choosing where to send your scores generally happens before test administration--that is, many students don't know whether or not they have something to "hide" until after the test, until after they've already made the decision about where to send scores. Second, students who apply to many places have to pay an extra fee to have scores sent (SAT includes a certain number of score reports with their testing fee--beyond those, you pay). That gives some students an incentive to NOT sent scores to schools which don't want them. Why spend the money?</p>
<p>Frankly, I think your point has some merits--among those who don't send scores will likely be students who don't test well. They don't send them because they know (or suspect) they're poor scores. A school who is willing to evaluate them on their other merits are attractive places to apply, indeed. So I believe you're right to assume that some of these students, had they submitted scores, would have been at the lower end. But to assume that every student without scores has a lower score than those who submitted them? That just doesn't feel credible to me, particularly given what we know about test scores and how the SAT handles score reports (as discussed above).</p>
<p>We've been very patient in trying to understand your position, even as you've implied we're dolts for not getting it. LOL Surely there is no need to dissolve into sarcasm.</p>
<p>Finally, it's just not good statistics to assume these students are lowest. I can't think of a reason why a statistician would want to include missing data in a percentile, but if compelled they would probably do one of two things. Option one is to substitute the mean score. That would put all of these students in tne middle, thereby having little effect on the 25th-75th percentile. Another option would be to try to impute an SAT based on other known factors about the student (GPA, math courses taken, gender, income, etc). But that's still pretty speculative. In the end, I feel that the percentile information you would have would be no more reliable or informative than the standard practice (which is to calculate percentiles only on students who submitted scores, making it clear to observers that this leaves off some number of students). I don't think an institution which does this standard practice should be regarded with derision or suspicion. Better to take the percentiles with a grain of salt, particularly given that the institution itself has decided they are not a good indicator of applicant quality.</p>
<p>There are general increases in the number of applications, and hence general increases in the number of students rejected, but there's no reason to believe that quality of the student bodies has increased one iota as a result, there are lots of reasons to believe that the result is more students end up attending "the wrong school", and we know absolutely nothing about whether the number of individuals applying to a class of schools has gone up or down.</p>
<p>Use the Ivies as an example. 23,000 places. Within the Ivies, let's assume an 85% yield. So there are 27,000 acceptances. There are 150,000 applications. But (eliminating the total hail marys and foreign applicants who don't know what they are doing), there are likely no more than 50-60,000 discreet applicants. In other words, the "Ivy acceptance rate" may be close to 50%. And it doesn't change as the average number of applications submitted by an applicant increases. But what does happen is that the student who would be happier at Dartmouth ends up at Columbia, the one at Harvard at Princeton, the one at Cornell at Yale, etc. The schools gained absolutely nothing by the exercise except more work on the part of the admissions office, a little bit of application money, and the prestige that comes from rejecting more applicants (and make no mistake about it - almost no prestige comes from accepting anyone, but lots from rejections.)</p>
<p>hoedown - Great post. You said it all.</p>