Of course, reality is that there no clean separation here.
Earning a high GPA commonly involves some combination of conscientiousness and learning aptitude in the ideal case, but some students try to game the system (looking for “easy A” courses/teachers) instead.
In terms of the SAT, your statement presumes that (a) the SAT measures all parts of learning aptitude that are of interest, (b) environmental influences do not affect learning aptitude, (c) environmental influences do not affect the accuracy of SAT’s measure of learning aptitude, and (d) there is no conscientiousness effect on the SAT. The existence of effective test preparation suggests that (b), (c), and (d) are not completely true.
The first 2 sentences of the article about why Americans don’t major in STEM are:
This is an extremely misleading quote, as it is choosing years that capture the dot com crash and the later 2007-08 stock crash, but do not include the following decade of tech prosperity. It you look at the past decade from 2009+, the results are completely different. Specific numbers are below. In all categories of STEM degrees, STEM degrees are increasing at a far faster rate than the overall average. CS is increasing at an especially high rate. However, when tech jobs become scarce or lower relative salary, such as during the dot.com crash, there is a very different pattern in the following years.
From 2009-10 to 2016-17 (most recent year listed in NCES)
CS Bachelors Degrees – Increased 80%
Engineering Degrees – Increased 59%
Math and Stats – Increased 50%
Biology – Increased 35%
Physical Sciences – Increased 34%
Humanities – Decreased 7%
Social Sciences – Decreased 8%
Education – Decreased 18%
The rate of STEM increase is becoming a problem at some selective colleges, particularly for CS. For example, CS has become Stanford’s most enrolled degree that has ever occurred in the 50+ years for which Stanford has captured degree enrollment. Some professors have expressed concerns about becoming “Stanford Institute of Technology.” CS has also become the most common degrees at some Ivies that have traditionally focused on liberal arts over vocationally focused fields, and is on the way to track to soon becoming the most common degree at most other Ivies where it is number 2. Specific numbers from Stanford are below:
Stanford CS Enrollment
2009 – 189 enrolled in CS
2010 – 247 enrolled in CS
2011 – 287 enrolled in CS
2012 – 360 enrolled in CS
2013 – 452 enrolled in CS
2014 – 574 enrolled in CS
2015 – 661 enrolled in CS
2016 – 663 enrolled in CS
2017 – 703 enrolled in CS
2018 – 739 enrolled in CS
I agree that the SAT has changed significantly over time and is moving more in the direction of a test of application and knowledge that can be prepped for… more like the SAT II subjects or ACT. However, it’s more than just a test instrumentation issue. For example, in my STEM classes in college, the general pattern was the class grade was primarily based on a series of short problem sets and a small number of exams. A few also had some kind of final project. Work is very different. It is more like a huge multi-year lab project that requires input from many different sources. Problem sets in college always have a clear and straightforward solution, which can usually be obtained in a few minutes by using a basic application of the topic discussed in lecture/text. Problems at work rarely have simple solution in a lecture, text, or anything you were exposed to previously. Many problems may not have a good solution at all. Sometimes you figure out what the problem is yourself, instead of having it assigned. Communication with others and interacting as part of the team is far more important than in college. In many jobs, your performance may have as much or more to do with the performance of the team than your individual performance. Performance may be graded on a metric that wasn’t told to you beforehand or that has variety of biases. I could go on.
Older studies from before more recent SAT changes also showed reasonably similar results to current ones in terms of how well SAT predicts college academic performance. In studies of older SAT formats, the writing section of the SAT is generally the most predictive section for college academic performance, which fits with SAT II generally being more predictive than SAT I (example study is at https://cloudfront.escholarship.org/dist/prd/content/qt7306z0zf/qt7306z0zf.pdf ). The previously linked consciousness study found that HS GPA for more predictive of college GPA than was consciousness, and consciousness added a good amount to the college GPA prediction beyond HS GPA… more than did SAT I. While consciousness had some degree of correlation with HS GPA, both metrics were measuring a variety of different things .
Fitting more with this thread, the degree changes by race are below for highly selective non-LAC private colleges are below. I am defining “highly selective” as the 22 colleges for which admit rate <= 20% and 25th percentile ACT >= 30, as listed in NCES. I started in 2010 because this was the first year of the new federal racial definitgions and ended 6 years later in 2016-17 because that is the latest year in the NCES. The changes are expressed as a ratio, rather than percentage points. For example, Asian CS percent increased from 3.0% of Asians students majoring in CS in 2010-11 to 11.7% of Asian students majoring in CS in 2016-17. 11.7/3.0 = 3.91 = 291% increase.
Changes in % of Race Receiving Degrees Between 2010-11 and 2016-17
Computer Science – Black increased 205%, White increased 163%, Asian increased 291%
Math & Stats – Black increased 62%*, White increased 33%, Asian increased 43%
Physical Sciences – Black increased 39%, White decreased 2%, Asian no change
Biology – Black increased 35%, White increased 18%, Asian decreased 2%
Engineering – Black increased 22%, White increased 11%, Asian decreased 1%
*Only 2 colleges had more than 3 Black math or stats degrees – Harvard and Brown in 2016, and Harvard in 2010. ~Half of colleges had no Black students receiving math or stats degrees.
Percent STEM + Economics in 10-11: 27% Black, 36% White, 55% Asian
Percent STEM + Economics in 16-17: 38% Black, 45% White, 65% Asian
I find it interesting that you, @ChangeTheGame , and I have similiar opinions even though racially and culturally we have little in common. I did grow up poor though.
Anderson once said that he did not have a high opinion of most liberal arts with the exception of a few. Critical theory and post modern thought must be anathema to his focus on analytical writing. Being a philosophy major before attending Harvard Law, he must know about the Sokol Hoax, where a mathematical physicist was able to plant a Trojan Horse in a respected journal. It was utter gibberish, but it was published. Recently, three academics decided to “repeat” the experiment:
I am glad, at least to see that the more rigourous disciplines did not fall for such nonsense. I also hope most URM do not end up with these majors. It may be acceptable in academia, but I can not see employers look kindly at them.
@Canuckguy Which liberal arts do you not have a high opinion of? Math/Statistics, CS, the physical and biological sciences, literature, history, philosophy, foreign languages, or just the social sciences? I mean, “liberal arts” covers pretty much everything except engineering and business and vocational majors
Edited to add I see you said ANDERSON has that opinion, not you.
But I am curious as to what are considered to be " the more rigourous disciplines" that you hope URM don’t wind up in, and why?
The question that I have is why do Law/Business/Med schools do this? I think most of us would agree that the MIT grad with a B average in STEM is probably more capable (… dare I say more intelligent?) than the soft major grad with an A average, especially since an A average may represent the median grade for that major, and many STEM grads at top universities often have very high gpa’s in non-STEM courses.
Certainly admissions committees in Law/Business/Med are aware of this? One possible reason for this behavior may be the pernicious effect of US News-type of ranking lists on professional schools and the desire of a school to have enrollees with high undergraduate GPAs.
A second possible cynical reason for turning a blind eye is because it provides a pathway for professional school admissions committees to rationalize certain aspects of “holistic” admissions—a way to improve the “diversity” of their enrolled students while being able to virtue signal at the same time (… our numbers for X group are up and by the way the average gpa is still high!) What goes unsaid, is that many with high GPAs are students majoring in Sustainable Development in Latin America, Anthropology as Human Rights Activism; Global Justice, etc—stereotypical examples given in the article above with minimal intellectual content. This way a middling MCAT/GRE/LSAT score can be paired with a high GPA to help construct an admitted class that artificially meets desired metrics.
Since law school ranking is (a) a big deal in law employment (law schools want their graduates employed), and (b) very stats (LSAT and GPA) sensitive, it is no surprise that law school admissions is very stats (LSAT and GPA) sensitive, with undergraduate major and course selection mattering relatively little. See http://schools.lawschoolnumbers.com/ for some law school admissions scatterplots.
Regarding difficulty of majors, not everyone can do well as an English major. It is likely that a significant percentage of MIT students would have trouble earning a top-14-law-school-worthy GPA as an English major.
Medical school admissions is rather different, but (high) stats (MCAT and GPA) are typically a first cutoff before the more holistic review, interviews, and such are reached.
Absolutely do not agree. For law school? Analyzing text and effective writing is critical. There’s a reason undergrads interested in law often choose English or Philosophy or History or Poli Sci or another social science and it’s not because they think it gets them an better GPA. It’s because they are great prep for law school and a career in law.
And honestly, “the NYU lit grad went to Harvard and the MIT grad was wait-listed at American” - did they both apply to Harvard and American? What else did they bring to the table besides their major and GPA? What was the MIT grad’s major - also English or something else? Did you know some incredibly bright accomplished students go to NYU?
First of all, the author seems ignorant of the fact that all of STEM except “E” (and “T” if not CS) are in fact liberal arts subjects. That aside, many people study something non-STEM because that is what they enjoy, are good at or furthers their career goals. You want to work for the state dept? You better have a foreign language, or several. You want to be a lawyer? You better know how to analyze text and write well. And so on.
Maybe the law school admissions folks read the MIT grad’s application essay?
Law is all about writing and communication skills. Something that doesn’t tend to be emphasized or developed in STEM courses.
But then again, that quote doesn’t tell us anything at all about the MIT student’s LSAT score. The assumption is simply made that the MIT STEM student must be smarter than the NYU English major. Why? I think it’s quite likely that the STEM major might have flubbed the LSAT, which is testing a very different type of thought process.
Most students choose their majors based on their interests, not on the perceived level of difficulty. But then again, how to define difficulty? I’m sure there are English majors who would do quite well in STEM courses, but simply have little interest in pursuing a career in sciences, but love the stimulation they get from reading broadly and the opportunity for creative expression. And STEM majors who would struggle with the writing expectations on humanities/social sciences side.
And no, I wouldn’t agree with the premise in any event. I’d assume that B student at MIT knew more about whatever it is that student majored in than the A student at NYU… but other than that, if I was in charge of admissions, an A average is better than a B average no matter what the major. And for law or business, I’d definitely place higher value on majors like English, history, or poli sci over most STEM majors. For one thing, at top law schools, the English or history major is likely to be much better prepared for the style of teaching – Socratic method and reading of case books.
Probably the B average was the main problem in getting into law schools ranked high enough for good law employment prospects.
MIT does require four communication-intensive courses (two in humanities / arts / social studies, and two in upper level major courses) for graduation, more than most schools do. Also, MIT’s general education requirements include about a fourth of the total bachelor’s degree course work in humanities, arts, and social studies.
STEM majors in general do not seem to do poorly on the LSAT, according to https://www.lsac.org/sites/default/files/media/2017-18_applicants-major_0.pdf . Math, statistics, and physics majors appear to do especially well, probably because of the logic puzzle section. Of course, that does not necessarily mean that the specific B average MIT student did well on the LSAT.
MIT’s average GPA was 3.39 in 2015, according to http://www.gradeinflation.com/MIT.html , so the B average student was in the bottom half of MIT students.
It does look like Kenneth Anderson, who wrote http://volokh.com/2011/11/09/reforming-higher-education-incentives-stem-majors-and-liberal-arts-majors-the-education-versus-credential-tradeoff/ is saying that both high school performance (in terms of getting the student into a university like MIT) and major (based on his opinion of “hard” versus “easy” majors) are more important than how well the student does in college courses. While pre-law students do try to game the GPA-focused law school admissions (by looking for “easy A” courses, as he mentions), doing well in college should count for more than high school performance when it comes to professional schools.
The highest LSAT scores are candidates that majored in Math/statistics/physics. Often these folks have higher VERBAL GRE scores that candidates in non-STEM majors.
Also, there is a difference between an English major and some of the light-weight, non-STEM majors that the article I was quoting was making fun of. For example the lowest average LSAT scores are in candidates that majored in Business Education, Home Economics, Police Administration, Indian Studies, etc.
The point of the article is that professional schools don’t care if you major in Indian Studies as long as you get a high GPA.
Also I get that liberal arts colleges often have science/math/CS majors, but I think that it’s common to use the term “liberal arts” as shorthand to reference the nonSTEM majors. Similar to how folks might ask if “Ivy league” schools might have a preference for A or B. Of course, on CC we’ll often get a retort about how “Ivy league” shouldn’t be used as a synonym for high caliber schools (… don’t you know that the Ivy league is just an athletic league… don’t you know that Stanford/MIT isn’t part of the ivy league, etc.), but as long as the implied meaning of the term is quite clear, it doesn’t bother me and I don’t think that using the terms in this manner always implies ignorance.
I don’t have the time to research this, but I’m relatively sure that I can find studies demonstrating that there is a significantly higher percentage of physicists/statisticians/mathematicians who can write well compared with english majors who can compete in physics/statistics/math.
Common, but incorrect, since the S (science) and M (math) in STEM are part of liberal arts. It is the T ((engineering) technology, a less common group of majors) and E (engineering) that are not part of liberal arts.
If you want to specifically refer to humanities and social studies, write “humanities and social studies”.
But note that liberal arts also does not include majors like business, home economics, police administration, and other similar (i.e. explicitly pre-professional) majors.
Some of the STEM majors had high LSATs, particularly Mathematics, but 3 of the top 4 were also non-STEM. With the exception of math, there was also a general GPA difference between STEM and non-STEM. Math tied with Classics for the highest GPA, while the other 3 STEM had the lowest GPAs among the top 10. Among the 4 STEM in the top 10 highest LSAT scores, the median GPA was 3.37; while the median GPA among the 6 non-STEM in the top 10 LSAT was 3.56 – approximately 0.2 difference. Part of this difference may relate to engineering majors being more likely to attend public colleges, and public colleges tending to have lesser grade inflation than privates. However, within the same colleges STEM majors tend to average slightly lower GPAs than the overall average, particularly among science majors.
Art History – 157.4 LSAT, 3.41 GPA, 80% admit rate
International Studies – 157.3 LSAT, 3.58 GPA, 86% admit rate
Foreign Languages – 157.3 LSAT, 3.57 GPA, 85% admit rate
*Mathematics had 2nd highest LSAT of all majors, after Statistics. Math also had the 2nd highest GPA of all majors.
What’s the point of comparing STEM majors with non-STEM majors when it comes to law school admission? Few top STEM students apply to go to law school. If they did, the result would be significantly different. The underlying applicant pool really matters.
The claim was at MIT, the B STEM student who applies to law school was more capable/“intelligent” than the “soft major” grad with an A average due to grading differences between subjects. If you look at a constant standardized LSAT score, the difference between GPA between STEM and “soft major” is nothing approaching the difference between a B and A GPA.
For example, applicants from the following majors all have an average LSAT score of ~156. This group includes chemistry, which one study found was the most harshly graded major. The 3 STEM majors do indeed have a lower average GPA among applicants with similar average LSAT score from the non-STEM (“soft majors”). However, that GPA difference between the STEM and “soft majors” is only ~0.1 on a 4.0 GPA scale – far less than the difference between a 4.0 (A) and 3.0 (B) GPA.
GPA Differences Among Similar Average LSAT Score Majors
Music – 3.54
Religious Studies – 3.52
Women’s Studies – 3.51
Interdisciplinary Studies – 3.49
History – 3.47
Regarding MIT GPAs specifically, MIT uses a unique grading system with a 5.0 maximum. It 2017, the average GPA among MIT fraternities was 4.48 (https://studentlife.mit.edu/sites/default/files/FSILG%20Spring%202017%20Grade%20Report.pdf ). All fraternities at MIT averaged above a 4.0. I expect fraternities to average lower grades than the general student body for a variety of reasons, so I’d expect most MIT students average well above 4.5, and I’d expect a B MIT student to far below typical , regardless of major. I don’t think you can make any kind of an assumption about a B student in STEM being more capable/“intelligent” than a A student in a “soft major,.”
In the article the friend “…was astonished and troubled to find that the MIT grad with the B average in STEM had fared far less well than the NYU English major.”
I don’t think that he would be “astonished and troubled” if the MIT grad was someone far below typical. This article was written in 2011.
Also note that the MIT grad had a B average in STEM courses. I would guess that his overall GPA in all courses including nonSTEM classes would be higher (and perhaps 4.2 or higher). Higher grades in nonSTEM courses is often the case for premeds at MIT and I’m guessing this is also the case for prelaw students.
Hence, again, I don’t think that the MIT grad with a B average in STEM courses is someone far below typical. Also given the well known grade inflation for humanities majors, I would guess that someone with an A average in english may be rather typical as well.
So 2 graduates who are within the typical range for each school. I actually know a number of MIT grads with mostly Bs in STEM classes and a number of other grads from other schools with mostly As in english. Are folks truly saying that a typical MIT student with Bs in STEM courses is NOT more capable than an NYU student with As in english? If so, so be it, but I would beg to differ. In my experience, the typical (or even below typical) student that matriculates at MIT has better high school academic credentials (including in nonSTEM courses) than the above typical student that matriculates at NYU.
To Calmom: “Most students choose their majors based on their interests, not on the perceived level of difficulty.”
A chart in this reference: https://www.businessinsider.com/science-majors-and-grade-inflation-2013-7
shows how likely students think it is that they’ll major in any given field (bottom) versus the actual likelihood that they’ll end up sticking with it. There’s a huge gap between initial perceptions and reality for science. In other words, many students (particularly in STEM) may start off with a major based on their interests, but in the end, they often graduate with a new major that has a reduced level of difficulty.
“The highest LSAT scores are candidates that majored in Math/statistics/physics. Often these folks have higher VERBAL GRE scores that candidates in non-STEM majors.”
That’s just a handful of people taking the test compared to the more popular majors. And why are these stem majors, who are supposed to be in highly employable fields taking the LSAT anyway? Shouldn’t they be off changing the world.
While MIT uses a 1-5 GPA scale, B grades convert to 3.00 for law school purposes using the more common 0-4 GPA scale. So the MIT sophomore average of 4.2 in 2007 converts to 3.2 for comparison to other schools using the more common 0-4 GPA scale.
humanities +0.3
engineering +0.2
social studies +0.2
But note that NYU in 2007 was heavily arts and social studies majors, while MIT was mostly a mix of science and engineering majors, so the difference in grade inflation by major are already baked into their overall GPA averages.
So do you judge college students and college graduates by their high school credentials and their family finances (which are the factors most influential in the names of their colleges) more than how well they actually do in college?
Grade inflation causes average grades over time, and upperclassmen tend to have a higher GPAs than sophomores (the first year in which students are graded on a non P/F scale for all semesters). For example, as mentioned in the link from my previous post, the average MIT fraternity GPA was 4.48 in 2017. In 2009, the average fraternity GPA was 4.26. The average GPA increased by 0.22 over a span of 8 years for a grade inflation rate of 0.0275 by year. If we assume the same grade inflation rate for all MIT students, then that suggests a sophomore GPA of 4.2 GPA in 2007 + 12 years * 0.0275/year = 4.53 today. If we consider upperclassmen tend to have higher GPA than underclassmen and half of freshman year is pass/fail, then that should increase the average GPA to be even higher.
The context of the quote with discussion about GPA by major implies cumm. GPA not course GPA – the B average was the STEM major’s law school app GPA, and the A average was the English major’s law school GPA. Neither was just the GPA in STEM courses or just the GPA in English courses.
Below typical for MIT students. At selective colleges, GPA distribution tends to be clustered near the mean and drop off quickly at the lower end, particularly below B, as such colleges tend to give few C’s. For example, in Harvard’s senior survey >99.5% of students reported GPA’s of 3.0+. However, MIT does not publish GPAs by major, so we have no way to confirm how many have a B average by major. We only know, a B average is well below the overall average at MIT or nearly any other highly selective college.
If you mean at NYU, their Latin cutoffs are at https://www.nyu.edu/students/student-information-and-resources/registration-records-and-graduation/graduation-and-diplomas/graduation-honors.html . The Latin cutoff for top 5% GPA was below 4.0 at every school, suggesting 4.0 GPAs are not common, but a few students do reach 3.9. If you mean A average as >3.5 GPA, then I’d agree that it’s common. At MIT English majors obviously aren’t common. There was one kid who majored in creative writing last year, which is part of a science/engineering interdisciplinary program… too small a sample size to estimate much of anything.
There is too little information to know who is more or less capable, especially without even knowing LSAT scores or any other aspect of the law school application. However, it’s certainly possible for excellent students to attend colleges besides HYPSM…, such as NYU. And many of those excellent students do not choose STEM majors, particularly among those who are planning to attend law school. There are also many students at MIT who would not make good attorneys. It’s not just a given students who get GPAs at the lower end of the class at MIT would be better attorneys than A students at NYU.
We’re not talking about averages here – we’re looking at an anecdote of an A student at one college NYU) vs. a B student from another (MIT) with the assertion that the B student must be smarter because he or she attended a college that was statistically harder to get into — without knowing anything else about either student. Which is a decidedly unscientific way to come to any conclusions at all about which student might be smarter or more capable.
There are some students at NYU who entered with higher stats than some students at MIT. NYU is a dream school for a lot of students, and actually can be very generous with financial aid for students at the very top of their applicant pool. And down the line, I am sure there are many NYU graduates who score higher on tests like LSAT, MCAT, GRE, or GMAT than many MIT grads, whatever their high school stats may have been. Because that’s just the way the way that things work. Just as there are many women who are taller than many men, even though it is indisputable that the average man is 5 inches taller than the average woman, and that the tallest men are much.
