Welcome to College Confidential!

The leading college-bound community on the web

Sign Up For Free

Join for FREE, and start talking with other members, weighing in on community discussions, and more.

Also, by registering and logging in you'll see fewer ads and pesky welcome messages (like this one!)

As a CC member, you can:

  • Reply to threads, and start your own.
  • Post reviews of your campus visits.
  • Find hundreds of pages of informative articles.
  • Search from over 3 million scholarships.
Please take a moment to read our updated TOS, Privacy Policy, and Forum Rules.

Differences in SAT ranges at top schools

collegehelpcollegehelp Registered User Posts: 6,523 Senior Member
Help me understand this. Caltech and Harvard both have Math + CR 75th percentiles of 1600 but the 25th percentile at Caltech is 1500 and the 25th percentile at Harvard is 1400. The 75th percentile is the same but there is a 100 point difference in 25th percentiles. Why? Does this mean Harvard relies more on "intangibles" for admissions decisions? That's a pretty big difference. The interquartile range at Harvard is 200 while the interquartile range at Caltech is 100.

Another example would be Dartmouth versus Tufts. They both have midpoints of 1445 (2015 data, I think) but Dartmouth has an interquartile range of 230 while tufts has an interquartile range of 150. That's an 80 point difference. I would think that Dartmouth and Tufts would be more similar in applicant pool than Harvard versus Caltech but there is still a big difference in SAT range.

Tulane has an interquartile range of 170 and UCLA has an interquartile range of 290. This is a 120 point difference. Yet they have the same SAT midpoint (1325).

What is going on? Do these big disparities in SAT ranges (despite identical midpoints) mean anything about the admissions process or applicant pool?

Why do schools with similar average SATs differ wildly in SAT variability among the student body?
«1

Replies to: Differences in SAT ranges at top schools

  • billcshobillcsho Registered User Posts: 16,479 Senior Member
    When the 75th percentile hit the maximum value, you know it is far from a normal distribution. In addition, Harvard admits a lot more students than Caltech.
  • DadTwoGirlsDadTwoGirls Registered User Posts: 2,318 Senior Member
    It depends upon how the university decides to admit students. Harvard almost certainly could have a 25th percentile SAT of 1500 if they wanted to admit people based on SAT score. Presumably they are looking at something else. This shouldn't be a surprise to anyone by now.
  • ucbalumnusucbalumnus Registered User Posts: 62,122 Senior Member
    edited May 16
    Why do schools with similar average SATs differ wildly in SAT variability among the student body?

    There can be any number of reasons.

    a. The relative weighting of test scores versus other criteria varies. For example, UCLA (and other UCs) have a lower relative weighting of test scores relative to HS GPA than most other schools of similar selectivity. Heavier relative weighting of subjectively graded criteria like essays and recommendations that is non-observable from outside the admissions office can result in the more observable criteria like test scores and HS GPA having a wider distribution.

    b. Some schools admit by division or major, and different divisions or majors may be very different in selectivity, resulting in a wider scattering of test scores and HS GPA across the school. So entire school stats may not be very useful, unless there is one division that makes up most of the school and the applicant is looking at that division.

    c. Some schools use "level of applicant's interest" to reject or waitlist "overqualified" applicants. Obviously, this causes them to lose some top end students who may have otherwise attended, shrinking their stat ranges at the top end.

    All of the above mean that using test scores as a proxy of admission selectivity may be misleading in many cases.
  • collegehelpcollegehelp Registered User Posts: 6,523 Senior Member
    Insightful analysis from ucbalumnis. I understand your points.

    Focusing on the SAT factor alone for the moment, does it mean, among schools with "equal" average selectivity (in the SAT sense), that schools with wider SAT ranges are actually easier to get into? To really understand selectivity (in the SAT sense) you have to consider the midpoint and the interquartile range?
  • ucbalumnusucbalumnus Registered User Posts: 62,122 Senior Member
    Focusing on the SAT factor alone for the moment, does it mean, among schools with "equal" average selectivity (in the SAT sense), that schools with wider SAT ranges are actually easier to get into?

    Not necessarily. It depends on what other aspects the college considers, and what the applicant has.

    For example, a Texas resident student may see that the University of Texas at Austin has an SAT range that is probably around 1170 to 1390. Does this mean that it is easy to get into if the student has a 1550 SAT, or that the student has a reasonable chance even with an 1170 SAT?

    Actually, the information given using only test scores is insufficient to answer the question, since the University of Texas at Austin admits most of its frosh with class rank as the sole criterion after meeting baseline criteria (such as high school course pattern). If the Texas resident student is in the top 7%, s/he can consider the campus an automatic admission safety (although not necessarily automatic for competitive majors). Outside of the top 7%, s/he must consider it a reach due to the small remaining number of spots, competing with the non-top-7% and all out-of-state applicants.

    Test scores may be a convenient way to compare admission selectivity across colleges because they are observable to outsiders and standardized, but they are often only a small part of the criteria used. So taking them as the definitive answer for a college's selectivity risks making incorrect conclusions.
  • collegehelpcollegehelp Registered User Posts: 6,523 Senior Member
    Thanks, ucbalumnus. I take your point and agree that SATs alone don't provide a complete picture. Still, I am surprised by the large differences in SAT variability among students at schools with more or less equivalent SAT averages. You have pointed out several possible explanations. Are there other possibilities? I would be surprised if the factors you named could account for such large differences in SAT spread among similar schools but I suppose it is possible.

    Are applicant pools less uniform at some schools? Is Harvard more likely to accept the B student with an Olympic medal?

    Sushiritto, your comment was hilarious! Made me laugh.
  • theloniusmonktheloniusmonk Registered User Posts: 821 Member
    This is a very interesting question, so I googled football stats for Harvard athletes and looks like their average SAT for M/CR is 1360 (on the old 2400 scaale) so figure that for 10-12% of athletes in the class, Harvard's average and 25% will be influenced by the 1360 while Cal Tech's will not. BTW 1360 is very good for an athlete. Similarly you can figure out the other ways the two schools are different, but as someone said above it will come down to the diversity (URM, legacy, first gen) of Harvard vs. the uniformity of Cal Tech.
  • avavidaavavida Registered User Posts: 6 New Member
    The most commonly used and accepted test taken by potential college students is the SAT Reasoning Test, which is distributed by the College Board. This test helps admissions officers to analyze a student’s ability to think critically and logically and to reason out the answers to problems. There are three major sections on the test, the writing section, the critical reading section and the mathematics section.
    https://conquercollege.quora.com/SAT-and-ACT-Information
  • YnotgoYnotgo Registered User Posts: 3,419 Senior Member
    edited May 19
    @collegehelp Regarding the score range for Harvard vs. Caltech, here are some reasons.

    -- Caltech admitted 525 aiming for a class of ~235 this year. Harvard admitted >2000, so ~10x as many.

    -- Caltech wants students who can handle the load. Course rigor at Caltech is pretty darn high. Because of the core requirements, there is no easy path to graduation that avoids a full year of math, full year of physics, and 2/3 year of chemistry. Plenty of theory in all that, and most majors require a lot more STEM. (Many entering Caltech students will have taken all of AP Calc BC, AP Physics C, AP Chemistry, AP Bio, and AP CS. Few will test out of any freshman courses and no AP credit is given.)

    -- Caltech wants students who will be a good "fit" for its quirky environment. The admissions essays ask questions that are very specific to this issue--3 STEM experiences, ethical challenge faced, quirky interest, diversity contribution, why Caltech, and submit research papers.

    -- Harvard's essay prompts are quite vague in comparison to Caltech's. Harvard wants a much more diverse class as far as interests. (And cares more about having a diverse class as far as background.)

    -- Caltech doesn't want to admit students who aren't capable of doing well. So, even athletes have to have the same scores as other students. (Or better -- after all sports take time.) Caltech sports teams are notable mainly for their long losing streaks.

    -- The Caltech students I've met are not the type to have prepped a lot for those scores. Many probably got those scores with one practice test and didn't view the SAT/ACT tests and STEM SAT Subject Tests as particularly difficult, especially math. Caltech also asks for AMC/AIME math scores, which are more challenging.
    Post edited by skieurope on
  • collegehelpcollegehelp Registered User Posts: 6,523 Senior Member
    edited May 19

    Seriously, thanks for the insights, Ynotgo. Any thoughts on the other schools I mentioned with comparable averages but disparate ranges? I keep thinking that, when schools are similar in average SAT, the one with the greater SAT range will be an easier school for admission. Put another way, schools with equivalent prestige may not be equivalent in selectivity in the SAT sense.
    Post edited by skieurope on
  • theloniusmonktheloniusmonk Registered User Posts: 821 Member
    the standard deviations on the sat midpoint has to come from the diversity of the class. Dartmouth has 48% white in the class of 2020, while Tufts has 67%. Students of color means URM and Asians, so the more non-white you have, the bigger std deviation you'll get and so the bigger 25-75 spread. Whether that helps you or not would depend on race, ethnicity, family background, as it does with any school.
  • YnotgoYnotgo Registered User Posts: 3,419 Senior Member
    @collegehelp Sorry, I'm from the West Coast. I just barely know where Dartmouth and Tulane are located, and wouldn't know where Tufts is without Googling.
  • roethlisburgerroethlisburger Registered User Posts: 1,574 Senior Member
    edited May 18
    First, I doubt Harvard's 75% is 1600 even with super scoring. I suspect you just added the 75% math to 75% CR, which isn't mathematically valid. Second, Caltech emphasizes academics, while Harvard is much more holistic in admissions.
  • skieuropeskieurope Super Moderator Posts: 25,862 Super Moderator
    Caltech and Harvard both have Math + CR 75th percentiles of 1600
    No they don't, or at least that is not what is reported on the CDS; you can't just add numbers together. They have 75th percentile of 800 for each section. The same student does not necessarily have an 800 on each. I'm sure that each college has plenty of people who scored 1600, but it's not 25% of the class, IMO.
  • ucbalumnusucbalumnus Registered User Posts: 62,122 Senior Member
    I keep thinking that, when schools are similar in average SAT, the one with the greater SAT range will be an easier school for admission. Put another way, schools with equivalent prestige may not be equivalent in selectivity in the SAT sense.

    Stop thinking that SAT scores are the only indicator of college admission selectivity worth caring about.
«1
Sign In or Register to comment.