The Smartest Colleges in America

College Search & Selection
1

<p>The</a> Smartest Colleges in America - Yahoo Finance</p>

<p>"Schools are constantly ranked on everything from the beauty of their campuses to their party scenes, but rarely on their students' intelligence. </p>

<p>Exclusively for Business Insider, Dr. Jonathan Wai, a Duke University Talent Identification Program research scientist and psychologist, computed a new ranking of America's colleges and universities based purely on smarts, as shown by the student body's average scores on standardized tests.</p>

<p>Though these tests are often criticized, research shows that both the SAT and ACT are excellent measures of general cognitive ability, and the scores of a school's student body give an accurate snapshot of their abilities. </p>

<p>Cognitive training company Lumosity took a crack at this last year, ranking more than 400 schools based on students' performance on brain-training games. There were a couple of problems with its list, however, as they only had data available from students who played the games, meaning some of the smartest schools got ignored, and there is no peer-reviewed research showing their games are a reliable and valid measure of smarts. </p>

<p>Now, for the first time, we present the real smartest colleges in America. (Read more about our methodology and reasoning here.)"</p>

<p>Thoughts?</p>

2

<p>Perhaps it should be titled “Colleges with frosh with the highest SAT CR+M and ACT scores”.</p>

<p>Basically, it is a selectivity-based ranking that focuses on one particular aspect of selectivity.</p>

3

<p>Well, well, well, look at little ol’ Boston College at #25. Who’da think it?! This MUST be s sham! …even beating out Notre Dame (34th) and that one in DC that always looks down its long, pointy nose at BC – Georgetown (122th!!!) --poor thing.</p>

4

<p>Hasn’t stateuniversity.com been doing pretty much the same thing for years?
[url=&lt;a href=“USA University College Directory - U.S. University Directory - State Universities and College Rankings”&gt;Top 500 Ranked Colleges - Highest SAT 75th Percentile Scores]College</a> Rankings - Top 500 Ranked Colleges - Highest SAT 75th Percentile Scores - StateUniversity.com<a href=“it%20uses%20only%20the%2075th%20percentile%20SAT%20scores%20to%20arrive%20at%20a%20similar%20list%20of%20top%2010%20schools”>/url</a></p>

5

<p>SAT or ACT scores do not measure cognitive ability or “g”. Mensa stopped using the SAT scores long ago. </p>

<p>There is a perfect correlation between SAT scores and family income.
<a href=“http://professionals.collegeboard.com/profdownload/cbs-2009-national-TOTAL-GROUP.pdf[/url]”>http://professionals.collegeboard.com/profdownload/cbs-2009-national-TOTAL-GROUP.pdf&lt;/a&gt; (Table11). It may indicate that higher income families may afford to pay for (i) expensive test prep classes, (ii) repeat tests that provides their children the opportunity to enhance their test scores. </p>

<p>Higher average test scores of incoming freshmen at these universities may indicate that the students may be better prepared for college, and not necessarily high cognitive ability.</p>

6

<p><a href=“it%20uses%20only%20the%2075th%20percentile%20SAT%20scores%20to%20arrive%20at%20a%20similar%20list%20of%20top%2010%20schools”>QUOTE=tk21769</a>

[/QUOTE]
</p>

<p>However, using 75th percentile may give a different list from using the average, due to differences in how “wide” or “narrow” the distribution of SAT scores is at each school.</p>

7

<p>I can say students with lower SAT test scores are not necessarily with lower cognitive ability. But for those who scored high, high cognitive ability is a necessity. The only question is how high of the cognitive ability is high enough. </p>

<p>My daughter is Stanford-Binet IQ test at 156 (99.98 percentile), she did well without any test prep help and is at Cornell University Engineering now. She graduated high school #1, and the #5 goes to Harvard. As she was always #1 of her classes, with hard work now she is not even top 10% at Cornell Engineering. Apparently there are more smart and talented young people at Cornell University than I thought. That was my motivation to come up with my research using the official common data set from universities which are usually viewed as top 20 schools. </p>

<p>After careful analysis, I have a list for Fall 2012-Spring 2013 enrolled freshman class.
(I started with Fall 2011-Spring 2012 analysis and had it in the link at the bottom of this post. I think it will be helpful for prospective students to set aside selectivity and look at the real student body. You can also find many common dataset data compiled all in one place there.)</p>

<p>Freshman student body SAT Strength 2012-2013
(Starting SAT scores of their top 800)</p>

<p>R#800 M#800 W#800 R+M</p>

<hr>

<p>740.00 780.00 -------- 1520 Cornell
757.19 755.75 761.47 1513 Harvard
744.99 766.49 756.49 1511 Penn
739.48 759.48 746.91 1499 Northwestern
737.85 757.85 -------- 1496 WUSTL
739.35 753.41 747.48 1493 Stanford
729.34 749.34 719.18 1479 Vanderbilt
735.29 739.70 -------- 1475 U.Chicago
732.01 735.60 738.81 1468 Yale
728.88 738.88 738.88 1468 Princeton
722.60 742.60 -------- 1465 U Notre Dame
729.41 733.09 733.09 1463 Columbia
711.91 741.91 727.67 1454 Duke
679.65 745.79 689.65 1425 MIT
706.04 710.64 720.64 1417 Brown
670.77 693.85 680.77 1365 J.Hopkins
674.71 684.28 684.71 1359 Dartmouth
650.50 693.89 670.50 1344 Emory </p>

<hr>

<p>Please read my explanations in the second half of post #1 in this following thread:
<a href=“http://talk.collegeconfidential.com/college-admissions/1530349-ivy-league-other-top-schools-student-body-academic-strength.html[/url]”>http://talk.collegeconfidential.com/college-admissions/1530349-ivy-league-other-top-schools-student-body-academic-strength.html&lt;/a&gt;&lt;/p&gt;

8

<p>Daswell, it’s nice that you did this list for the Ivies and other elite schools. Sticking to the list cited at the top of this thread, what I found interesting is that many of the schools within the top 100 have very bright students as well. It is a useful tool for those of us who want our kids to find academically challenging schools with peers of similar caliber but don’t want to use the subjective USNWR rankings as a starting point and aren’t looking at “status” schools.</p>

9

<p>

</p>

<p>Actually not true. Mensa will take older SAT scores, but they stopped taking more recent scores because test prep has made it something people can and do study for. But almost no one took it more than once back in the 70s and 80s, so they will still take those. Which implies that Mensa at least DOES think the SAT tests (or did test) cognitive ability, but also can be “jimmied” by studying.</p>

<p>One of my kid has a higher Stanford-Binet than 2Daswell’s, and is a freshman at one of the top 10 schools on that list. She loves it, says she is surrounded by brilliant students and professors. Workload is crazy heavy – but she could not be happier. “Smartest” college was at the top of her criteria list, and she got what she wanted.</p>

10

<p>

If your motivation is your daughter not being in the top 10%, wouldn’t it make more sense to look at the top 10% (or other top ##%) instead of the SAT score of the 800th worst student in the class? The 800th worst student in a freshman class of 1000 is in the bottom quarter. Cornell is the only college you compared where it would be in the top quarter, so it’s not surprising that Cornell did well with this methodology. If you want to compare 800 students in the class instead of the full class for some reason, why not at least estimate the average SAT for this group instead of the score of the 800th worst student?</p>

11

<p>I was surprised this thread hadn’t gotten more traction. I think the link in the original makes it hard to view the entire list without digging. Here’s the complete list.</p>

<p>[Complete</a> Ranking Of America’s Smartest Colleges - Business Insider](<a href=“http://www.businessinsider.com/complete-ranking-of-americas-smartest-colleges-2013-9]Complete”>Complete Ranking of America's Smartest Colleges)</p>

12

<p>Not sure if cognitive ability includes ability to (i) innovate & (ii) invent.</p>

<p>Just because someone studies harder for SAT & takes it multiple times to improve their scores, can we say that it is a measurement of one’s innate smartness? People who score well on standardized tests may also score high on cognitive tests, but not sure if there is a 1:1 correlation. </p>

<p>China’s Gaokao’s are supposed to be far tougher than SAT. [Gaokao:</a> Are Your Kids Smart Enough for China’s Toughest Test? (Tao Jones) - Speakeasy - WSJ](<a href=“Gaokao: Are Your Kids Smart Enough for China's Toughest Test? (Tao Jones) - WSJ”>Gaokao: Are Your Kids Smart Enough for China's Toughest Test? (Tao Jones) - WSJ) </p>

<p>I think all these standardized tests, including Gaokao, measure academic preparedness,and not necessarily cognitive ability or “g” & inventiveness.</p>

13

<p>(This forum stacks up quickly with new threads; I didn’t get back to this thread until now.)</p>

<p>“If your motivation is your daughter not being in the top 10%, wouldn’t it make more sense to look at the top 10% (or other top ##%) instead of the SAT score of the 800th worst student in the class?”</p>

<p>First, there is no data available for top 10% of any national universities. Then, the #5 I mentioned who goes to Harvard didn’t seem to have as much challenge. So I started to use common data set to figure out who has ‘more’ academically strong students. </p>

<p>Appearently you are still using percentage and didn’t catch my point. If you read the second part of post 1 in the link at the bottom of my previous post, you will get the idea. (or easier to click the following)
<a href=“http://talk.collegeconfidential.com/college-admissions/1530349-ivy-league-other-top-schools-student-body-academic-strength.html[/url]”>http://talk.collegeconfidential.com/college-admissions/1530349-ivy-league-other-top-schools-student-body-academic-strength.html&lt;/a&gt;&lt;/p&gt;

<p>Once a poster wrote this comment:
Yale freshman class has 76% scored above 700 on SAT Reading, Cornell has only 42%
Yale freshman class has 80% scored above 700 on SAT Math, Cornell has only 65%</p>

<p>And my answer:</p>

<p>“Reading
76% of Yale’s freshman class of 1355 is 1029
42% of Cornell’s freshman class of 3182 is 1336</p>

<p>Math
80% of Yale is freshman class of 1355 is 1084
65% of Cornell’s freshman class of 3182 is 2068</p>

<p>Cornell have more students scored 700 in CR, more students scored 700 in Math. For the entire Yale class Cornell can cover it once (with higher scores). Cornell is just bigger and from that point on can continue on to take more students.
“</p>

<p>It isn’t hard to understand how to use available data to see how strong a school’s student body is.</p>

<p>People need to think out of the box and not to take USNWR ranking as Bible because it is not, it is just a money-making tool. They change methodologies as they wanted to. If one paid attention, he/she should notice that 2014’s ranking is different from past several years when Harvard was at top. Why? They need that kind of change so people will buy their ranking books.</p>

<p>Read the last section of this page about their change of methodology and continue to read the second page:
[How</a> U.S. News Calculated the 2014 Best Colleges Rankings - US News and World Report](<a href=“http://www.usnews.com/education/best-colleges/articles/2013/09/09/how-us-news-calculated-the-2014-best-colleges-rankings?s_cid=related-links:TOP]How”>http://www.usnews.com/education/best-colleges/articles/2013/09/09/how-us-news-calculated-the-2014-best-colleges-rankings?s_cid=related-links:TOP)</p>

<p>Their methodology will continue to change and schools will be shuffled in the rankings. It is better that I analyze data myself and use rankings as references. </p>

<p>(OP’s ranking is again using percentage… it has something to it but I think other theory should be explored as well.)</p>

14

<p>Even high income folks have SAT scores on average far below what many here would call “high”. Barely a 1700 which would impress nobody here. It’s more like they are all above average Lake Wobegon style.</p>

15

<p>

If you mean SAT scores, some colleges publish relevant info. The selective colleges that do usually have all 800s in the top 10% and a good ways below that, making it an uninteresting figure. For example at Stanford, 25% of this year’s admitted class had an 800 on math and 24% had an 800 on verbal. In contrast at Brown, 17% of the class had an 800 on math and 18% on verbal. To get to the sub-800 range at the most selective colleges, you’d need to drop down well below 10%, such as top 25%, a stat which is readily available. </p>

<p>The point I was making is if your motivation was your daughter not being among the top 800 students, then sure, it would make sense to look at the top 800 students. However, you mentioned being in the top 10%. Yes, selective colleges with a large freshman size tend to have more top scoring students than colleges with a small freshman class size. However, I assume your reference to top % was in relation to grades. Grade curves are generally done such that a certain % of the class gets As, rather than reserve As for a specific number of students with no consideration to class size, so the % of class with … stats figure is still the relevant one. </p>

<p>Percentage is also the stat that I’d expect to be relevant to most students as it correlates with what type of students they are most likely to run into in their major, classes, dorms, etc. One exception would be if the college is large enough to support honors-type programs within the major. </p>

<p>

And I’d expect that Berkeley, Michigan, UVA, and similar get more 700+ scores than either college. In contrast Caltech would have a small fraction of the number of top scorers. Nevertheless, I’d expect that students as a whole are most likely to interact with top scoring students at Caltech.</p>

16

<p>^My motivation is my always-number-one, high-achieving, highly intelligent daughter getting 3.3 - 3.5 GPA (I had to constantly ask her to sleep more) and I didn’t know how strong Cornell’s student body really is by only glance at their admission data online, so I set out to figure it out. Then it grew bigger - to compare schools of similar caliber. I can see your concern that I didn’t stop at my initial intention and did go further to do some comparison between schools. But for a research, you don’t need to stay at your pre-research intention. It can grow or shrink or change direction depending on data and approach. I ended up not only to understand D’s school (by looking into their Engineering data) but also provide realistic information to those prospective students who are looking into top schools without information about which ones they fit better based on student body academic strength, size, and diversity of talents. I hope some have a clearer picture now.</p>

<p>Since you mentioned Berkeley, Michigan, UVA, let’s pick the one with best reputation and highest ranks in most systems: UC Berkeley. Cal enrolled 4,127 and Cornell enrolled 3,217 first time first year students. We can compare from 1,032 up to 2,412 student scores between UCB and Cornell. </p>

<p>For Berkeley and Cornell’s upper 1,032 freshman SAT scores:
(Enrolled freshman 2012-2013)</p>

<p>R#800 M#800 W#800 R+M</p>

<hr>

<p>725.84 764.42 --------- 1490.27 Cornell
719.98 769.98 749.98 1489.97 Cal</p>

<p>(similar for their upper 2,412 students with Cornell top Cal for less than one point)</p>

<p>Very close, Cornell the smaller school is a little higher on scores. Cal is a very strong school as well, but not what you expected like the comparison between Cornell and Yale where the bigger one can cover the smaller one entirely with stronger scores. And No, Cal’s number of students who scored 700+ on Math and Reading cannot cover the whole freshman class of Cornell students. </p>

<p>On ‘Stanford’s 25% admitted student scored 800 in Math’ part of your post- this is comparing apples and oranges. I am calculating using common data set data for ‘enrolled’ freshman, not ‘admitted’ students. If you look carefully, Stanford ‘admitted’ 2,423 students, that means 606 of them scored 800 in Math. But when you look at common data set, for 1,675 ‘enrolled’ students there are 441 (25%) scored 790 or higher; if only one of these students scored 790 and all others scored 800 you have 440 Stanford ‘enrolled’ freshmen scored 800. Cornell’s Math 75 percentile score is 780 which means 804 ‘enrolled’ students scored 780 or higher, it can be easily with 500 students scored 800. (Cornell’s Engineering school Math 75 percentile is 800, don’t know if 60 percentile is already 800) </p>

<p>By the way, at least 166 out of 606 Stanford admitted students whose Math score is 800 enrolled in other schools.</p>

<p>Engineering of top schools: (Now I understand this may be normal for the Engineering of this caliber)
<a href=“http://talk.collegeconfidential.com/cornell-university/1562337-cornell-parents-did-your-superachiever-fail-their-first-test.html[/url]”>http://talk.collegeconfidential.com/cornell-university/1562337-cornell-parents-did-your-superachiever-fail-their-first-test.html&lt;/a&gt;&lt;/p&gt;

<p>(I would like to do more on this college comparison thing and develop something truly useful for my grand children when I retire. :-).
For now, this is still the list I have:</p>

<p>Freshman student body SAT Strength 2012-2013
(Starting SAT scores of their top 800)</p>

<p>R#800 M#800 W#800 R+M</p>

<hr>

<p>740.00 780.00 -------- 1520 Cornell
757.19 755.75 761.47 1513 Harvard
744.99 766.49 756.49 1511 Penn
739.48 759.48 746.91 1499 Northwestern
737.85 757.85 -------- 1496 WUSTL
739.35 753.41 747.48 1493 Stanford
729.34 749.34 719.18 1479 Vanderbilt
735.29 739.70 -------- 1475 U.Chicago
732.01 735.60 738.81 1468 Yale
728.88 738.88 738.88 1468 Princeton
722.60 742.60 -------- 1465 U Notre Dame
729.41 733.09 733.09 1463 Columbia
711.91 741.91 727.67 1454 Duke
679.65 745.79 689.65 1425 MIT
706.04 710.64 720.64 1417 Brown
670.77 693.85 680.77 1365 J.Hopkins
674.71 684.28 684.71 1359 Dartmouth
650.50 693.89 670.50 1344 Emory </p>

<p>)</p>

17

<p>

Why are you looking at the test scores of the 1032nd worst student in the class when my comment mentioned more 700+ scores on the SAT? </p>

<p>The data in the 2012-2013 CDS shows Berkeley had approximately 4127 * 56% = 2311 700+ SAT scores on math and 4127 * 36% = 1486 700+ SAT scores on verbal. In contrast Cornell had approximately 3307 * 65% = 2150 700+ scores on math, and 3307 * 42% = 1389 700+ scores on verbal.*</p>

<p>So Berkeley comes out well ahead of Cornell on the 700+ SAT score measure I mentioned in my earlier post. Michigan has a larger freshman class than either Berkeley or Cornell, so it does much better than either school, as listed in the table below:</p>

<p>Approximate Number of 700+ SAT Scores*

  1. Michigan (6124 in freshman class) – 3240 Math, 1874 Verbal
  2. Berkeley (4127 in freshman class) – 2311 Math, 1486 Verbal
  3. Cornell (3307 in freshman class) – 2150 Math, 1389 Verbal</p>

<p>Note that the more students, the better ranking by this 700+ score method. For example, a Michigan alumni might try to show his school is “smartest” by looking at the 2500th worst student on math and 1500th worst student on verbal. By this measure, Michigan would come out far ahead of all the ivies, including Cornell, as well as any other selective college with significantly fewer students in their freshman class than Michigan.</p>

<p>*Calculating as if all students submitted SAT and none submitted ACT</p>

18

<p>“More on 700+” is not my methodology. I was just answering a question from others.</p>

<p>"Reading 700+
76% of Yale’s freshman class of 1355 is 1029
42% of Cornell’s freshman class of 3182 is 1356</p>

<p>Math 700+
80% of Yale’s freshman class of 1355 is 1084
65% of Cornell’s freshman class of 3182 is 2068
"</p>

<p>With above, I said Cornell’s number of students whose CR and Math scores are 700 and above can cover the entire Yale freshman class. Bigger Cal and Michigan, ,according to the numbers you gave, did not achieve the same result with Cornell. Bigger is not always stronger.</p>

<p>My methodology is to compare largest possible number of students within comparable range (800 students for 2012-2013) and it is totally achievable for the USNWR top 20 schools as I have already listed. (except Caltech and Rice which are too small to be compared with the others.)</p>

19

<p>The 800th worst student is an arbitrary selection that favors Cornell’s freshman class size of ~4x 800. If you choose either far more than 800 or far less than 800, the ranking varies dramatically. For example:</p>

<p>If I choose the 1600th worst student instead of the 800th worst, then I get the following results:
Approximate Scores of 1600th Worst Student in Class*

  1. University of Illinois – 800M, 690V
  2. University of Michigan – 760M, 700V
  3. Cal Berkeley – 740M, 690V
  4. Cornell – 750M, 650 V
  5. Stanford – 630M, 620 V
  6. Harvard – 610M, 590 V</p>

<p>Note that this order exactly follows the size of freshman class from largest to smallest. It’s not a coincidence.</p>

<p>And if I choose the 400th worst student instead of the 800th worst, then I get the following results:
Approximate Scores of 400th Worst Student in Class*

  1. Harvard – 800M, 790V
  2. Stanford – 800M, 790V
  3. Michigan – 800M, 780V
  4. Cal Berkeley – 800M, 770V
  5. University of Illinois – 800M, 770V
  6. Cornell – 800M, 770V</p>

<p>This time the smallest freshman class sizes of Harvard and Stanford come out on top, in spite of 400th worst being far lower percentile at H+S than at the large freshman class colleges. Looking at the 1600th worst student, H+S are last, yet looking at 400th worst student, H+S are first. Looking at the SAT scores of the worst arbitrary number x 100th student in the freshman class is a ridiculous way to measure “smartest colleges.”</p>

<p>*Again calculating as if all students submitted SAT and none submitted ACT. For 400th worst student, I am assuming 700+ Verbal percentage has a linear distribution for scores between 700 and 800. This should be treated as a rough estimate. All values are in tables are approximations rounded to the nearest multiple of 10</p>

20

<p>Do you mean 1600th best and 400th best, not 1600th worst and 400th worst?</p>

<p>The 400th student, counting from the bottom of the entering class, should have worse SAT scores than the 1600th student, counting from the bottom of the entering class.</p>