What Percentage of 1600s Get In?

Colleges and Universities A-Z Harvard University
41

<p>"what it's saying is that - on average - well over 90% of applicants who apply to Penn ED with SATs of 1600 who are in the top 10% of their class will be admitted. I contend that's actually pretty close to being spot on."</p>

<p>As a top 10% and 1600 scorer rejected (not deferred) from Penn ED, I may be merely wishing that your information is inaccurate. However, I suppose it is possible so I will grant you that your contention MAY be true. What puzzles me though is: why is Brown's rate so low, especially early decision? Also, at the meeting for Penn that I went to she told our group that over 1/2 of the Valedictorians and over 1/2 of the perfect SAT scorers were rejected - surely things haven't changed THAT dramatically in the past four years. Perhaps the survey data compiled over several years when admissions was easier? </p>

<p>Regardless of what impact this has on me I find the topic interesting and am anxious to hear what you have to say! :)</p>

42

<p>1/2 of the valedictorians and 1/2 of the 1600s being rejected could very easily fit into the stats presented. Did they tell you what percentage of valedictorians with 1600s they reject?</p>

43

<p>No they didn't. Oh I don't doubt Penn's statistics...it's just the one from this book were stated at a 97% acceptance rate, which seemed contradictory to what I heard at the Penn meeting.</p>

44

<p>Yes, I understand, and what I meant was that what Penn told you and what the book says can both be true, given that the vast majority (if not all) of the valedictorians they reject had low (or at least sub-1600) SAT scores, whereas the vast majority of the 1600s they reject were probably those not ranked too high.</p>

45

<p>yeah my rank wasn't particularly high but still top 10%. Did the book say Val + 1600 or just 1600, b/c I suppose 97% makes sense for 1600 Valedictorians...</p>

<p>EDIT: I do see your point more now though I kinda missed what you were saying. But if the book says top 10% + 1600 = 97% acceptance rate, I still should have fallen in that category.</p>

46

<p>I think that Brown's numbers were lower when we did our study in large part due to he fact that they had an Early Action program at the time. Our data showed that the marginal benefit of applying EA is less than that of ED. That makes sense if you see the schools as using ED as a means for imprining yield. I am very surprised to hear that you were not admitted to Penn in ED with a top 10% rank and perfect SATs. Did you take the most challenging curriculum? Were there any red flags in your file?</p>

47

<p>donmesw raises a very good point. The 50% of 1600s who are rejected may have poor grades, and the 50% of valedictorians may have low scores. Admissions offices are looking for distinction in both categories. My experience is that applicants who score high and who have weakness in their transcripts tend not to fare well. Schools tend to look for applicants who have "taken the best advantage of the opportunities afforded them."</p>

48

<p>The results presented in Table 5.2 are described as estimated chances of admission (as a function of SAT score) to each of a set of colleges based on regression analyses of your survey data. I think you have an unfortunately small number of cases for these analyses. You begin with only 3,200 cases and reduce this by 59% to ~1,300 by excluding females from your regressions. More importantly, you do not report the numbers in your sample who applied to individual colleges -- numbers that are by necessity VERY small. Given this, reporting the “SAT breakdown” chances at the college level is a highly questionable practice and leads to results that are simply not credible.</p>

<p>Let us look at a specific example regarding Harvard and MIT. For Regular Decision applicants with an SAT of 1400, you estimate a one tenth of one percent (0.1%) chance of admission to Harvard and a thirty-five plus (35.4%) chance of admission to MIT. I submit that those are bizarre numbers.</p>

49

<p>Eh. I got a 1600 and was top 1% of my class and got outright rejected by Stanford. Then again my essays weren't too great, and I neglected to put quite a few things for my ECs. Based on those numbers though, I should have quite a good chance RD at all my schools... But I'm more than a little sceptical.</p>

50

<p>
[quote]
Randomperson's comments seemed so odd to me, that I looked into it, re-reading the "Early Admissions Game", then quizzing one of the authors, who I happen to know.</p>

<p>In fact, the statistics for 1600-scorers shown in Table 5.2 on page 160 are for actual 1600 scorers, as shown in the survey data.</p>

<p>Randomperson may be confused by the fact that the study DID attempt to normalize for other factors (recruited athlete, legacy status, etc) in calculating the odds of admission, both early and regular, for applicants with the same SAT score.

[/quote]

My memory was wrong, but this statement is also incorrect. I checked the book out of the library, and the data on page 160 is absolutely not from actual 1600-scorers. Just read the explanation on page 159, which I copy verbatim:

[quote]
To refine our figures, we estimated the chances of admissions for a hypothetical male applicant who is approximately average within the survey in terms of activities and high school attended, and who has no other distinguishing characteristics. We assess the prospects of this average student four times, giving him four different sets of test scores varying from 650 to 800 on each of five SAT tests (the SAT-1 verbal and math, and three SAT-2 subject tests).

[/quote]

In other words, these are strictly estimated chances (which you would have realized had you read the caption right above the table). The authors took a male who was average in every sense outside test scores and endowed him with several different sets of scores, then estimating his chances. Unforunately, the results don't entirely reflect reality. There are quite a few applicants who are "average" in every way but test scores, and these are precisely the kinds of high-scoring applicants that elite schools reject (just take a look around this site).</p>

<p>My sense is that these estimated chances were meant to be demonstrative - meant to prove a point about the difference between early and regular admissions, not to pin down the chances of hypothetical applicants. Since the nuances of regression analysis are too cloudy for many people to appreciate, the authors also presented us with a more compelling chart with predicted chances. These numbers don't, however, give us an accurate perception of how 1600s fare.</p>

<p>And to Mr. Fairbanks: first, thank you for writing this book. It is a valuable expos</p>

51

<p>"These numbers don't, however, give us an accurate perception of how 1600s fare."</p>

<p>I suspect that they don't give us an accurate reading in many respects. The authors would have us believe that it is 354 times likelier for a 1400 scorer to be admitted RD to MIT than to Harvard.</p>

52

<p>
[quote]
I suspect that they don't give us an accurate reading in many respects. The authors would have us believe that it is 354 times likelier for a 1400 scorer to be admitted RD to MIT than to Harvard.

[/quote]

As I mentioned earlier, I don't think that was really the intent of the estimates. The problem is that they have been misused here and treated as solid statistics.</p>

53

<p>Unfortunately, I don't believe either of the prior two posters read, or fully understood, the comment made by one of the co-authors of the study earlier in this thread.</p>

54

<p>Yes, Byerly, I saw this:</p>

<p>
[quote]
I am one of the authors of the book, so perhaps I can clear up some of the confusion. We sourced our data from two sources. First, we got full access to the databases from 14 of the 20 most selective schools in the country. We were able to run regressions to show the effect of applying early action/decision while controlling for all relevant factors. But we agreed to protect the anonymity of the schools in exchange for access to the data. To validate our findings and to provide school-specific data, we then conducted surveys of several thousand seniors during their senior years. the students in the pool were selected from students in the top 10% of their high school class. As a result, the admission rates cited in the thread are higher than the rates for the pool of 1600's as a whole, because these applicants were also top classroom performers. Hope that helps to explain the difference.

[/quote]

First of all, that doesn't explain the issue with Brown that I discussed, which is illustrative of the general problems with these numbers. Secondly, you're missing my main point: these are estimated chances, not actual statistics like you initially claimed.</p>

55

<p>Statistics are estimates by definition.</p>

56

<p>I said they were actual statistics with respect to THE SURVEY - which is a fact the author of the study has corroborated. You must learn to read more closely.</p>

<p>A direct quote from my prior post:</p>

<p>"In fact, the statistics for 1600-scorers shown in Table 5.2 on page 160 are for actual 1600 scorers, as shown in the survey data."</p>

<p>You should obtain and read a copy of "The Early Admissions Game" in its entirety, rather than just riffing through it at the library.</p>

57

<p>I know fully what I'm talking about, Byerly, and I look forward to Mr. Fairbanks' response:</p>

<p>"Let us look at a specific example regarding Harvard and MIT. For Regular Decision applicants with an SAT of 1400, you estimate a one tenth of one percent (0.1%) chance of admission to Harvard and a thirty-five plus (35.4%) chance of admission to MIT. I submit that those are bizarre numbers."</p>

<p>Table 5.2, p. 160</p>

<p>And I don't care if the table is derived from regressions (as the book says) or serves to report actual data (as you seem to think), it is still obviously erroneous and misleading.</p>

58

<p>I don't think its "erroneous and misleading" in the slightest. We can just agree to disagree about that.</p>

<p>As for the "example" you posit, you are apparently simply unwilling to accept survey numbers you don't like from the year in question.</p>

59

<p>I'm saying that (at best) the survey is defective, probably due to its small sample size and unrepresentative nature. It has not been the case (as Table 5.2 reports) for <em>any</em> year that Harvard has admitted virtually no male applicants in the top 10% of their class and with an SAT of 1400 or lower. In fact, many Harvard students scored 1400 or lower. That's true now and was also true when this survey was administered. Harvard's 2004 25-75 range was 1400-1580.</p>

60

<p>You need to try harder (though you are clearly unwilling to do so) to understand the composition of the survey group vs total data for the class in question in that particular year.</p>