But my point is that picking a major that one is not well-suited for because of “ROI” doesn’t always work out so well. That’s the pitfall of choosing majors by perceived post-college employability. No one ever chose to major in philosophy for the money.
I think you are overestimating how much of grade performance is explained by SAT score, and underestimating how much support is often available to students with weaker math/science backgrounds at more selective colleges. The latter often includes multiple levels of intro math/science classes, with a placement test and placement officers to help assist in choosing the course appropriate for the student’s HS background.
Regarding SAT score GPA connection, different studies arrive at different specific % variance explained depending on methodology, but every study I have ever seen that reviews this found that SAT alone only explains a small minority of variance in college GPA. For example, the study at https://journals.sagepub.com/doi/full/10.1177/2332858416670601 found that SAT score only explained 14% of variance in first year GPA among 10k students in the CUNY system and 16% of variance in first year GPA among 10k kids in the Kentucky public system. ~85% of variance in grades depended on other factors besides SAT.
While SAT is correlated with GPA that doesn’t necessarily extend to GPA expectations based on whether the student’s SAT score is matched or mismatched with the average SAT score of the student body at the school he/she attends. Kids who are poorly prepared for college often do poorly at whatever college they attended – both more selective ones and less selective ones, regardless of whether their SAT score is mismatched in the up or down direction. It’s not obvious (to me) that the 1200 SAT kid would necessarily have a better chance of college success at a directional state with average SAT of 1000 than at Stanford, with an average SAT of approaching 1500.
I agree that the kid is likely to rank lower within his class at Stanford than at the less selective college, but the two colleges are likely to have completely different grade distributions, such that a rank within a particular class corresponds to a completely different grade. Highly selective private colleges typically have median GPAs in the A range. For example, the most recent Harvard senior survey lists a median cumulative GPA of 3.8, with 90% of the senior class at A or low A- cumulative average (>= 3.5). In contrast, many less selective colleges have median grades in the C range, with the majority not making it to graduation, and very few graduating with an A average. Suppose a particular directional state has 15% of kids graduating with A or low A average. Is it easier to be at among this top ~15% rank at the directional state with a low A or better average? Or is it easier to avoid being among the bottom ~10% at Harvard with a low A or better average?
Whether a particular student is encouraged/discouraged by the how he compares with his classmates varies quite a bit between different students and different colleges For example, does the college offer multiple levels of rigor for the intro math/science (pre-med) classes and help students choose the course sequence that is appropriate for their HS math/science background? Does the college offer support networks for students who are struggling (tutoring/office hours), and will the student who is struggling take advantage of them? Does the college encourage a collaborative atmosphere?.. or a competitive one? Is the student the type who struggles when he is not among the best… or when he is bored/not challenged? Is he the type who takes on characteristics of friends/colleagues that he works with, or is inspired by them to excel? There are far too many variables to make a simple call. Instead it depends on both the student and the college.
“So I’d guess that if the other 60% of a student’s grade is low in this class, then students drop and aren’t included in the grade guides.”
These are required classes right, so dropping them would put you behind a little, you could make up in the summer but if you need to pass Calc 1 in the fall to take Calc 2 in the spring, you have to pretty much take the grade. You don’t want to be messing with your math sequence at a place like Berkeley and majoring in CS or engineering, or chemistry sequence if you’re pre-med.
I forget the exact %, but in terms of premed, there’s a HUGE culling. I have a relative who was premed until Orgo and then graduated as an economics major. My kid at UMich has a friend, majoring in CS, who was failing Calc 3 (MV Calc), and now isn’t a CS major. These kids withdraw and switch majors and those grades aren’t included in grade guides.
And often taking the course a 2nd time will yield a better outcome. It did for me a 100 years ago.
Odd. Seems like s/he could have switched to LSA CS that does not require multivariable calculus that CSE (in the engineering division) requires. Perhaps there is more to the story, like struggling to barely get C grades in other math and CS courses.
I don’t think it’s necessarily odd, since we were told during orientation that 75-85% of freshman change majors at least once during their time at UMich. I dunno the entire story, but the kid had a 28 in the class at the time of withdrawal. From what I understand, there’s a difference in the level of prestige between the two degrees too.
I think you are working off theory and not necessarily reality. I’m working off the reality of what I’ve seen in 20 years of working at an average public high school. The pre-med wannabe students who are most successful getting into med school are those who attend a college where they are in the Top 25% of incoming students for both SAT and GPA. Yet when they attend a college where they are in the super tippy top (often due to finances) they return very wistful about the lack of peers and sometimes opportunities compared to others who chose better schools, so that should be considered if they can afford to. Note - they still make it in to med school. Note too that I’m talking about schools where they are in the top 1 or 2% of students - very few academic peers out there. This is not the case with most state flagships. It can just as easily be a private school as a state directional school.
It’s not our poorer (academic) students who head toward pre-med. When they head to college pretty much all of them graduate barring a medical issue or similar. But right off hand I can’t think of any* who went to a school where they were in the “middle of the pack” who still made it into med school. The vast, vast majority changed their plans having decided they weren’t of high enough caliber - many telling me they figured it out in their first science or math class.
- There are exceptions for the very talented where they'd be "middle of the pack" at Top 20 schools.
Your statement about poorly prepared students not doing well in college is why there’s the grade discrepancy between schools that you continue to relate. Those students don’t make it in to top schools. It really isn’t part of what I’m talking about with pre-med students. I’m talking about “relative” preparation. A 1200 student can still do really well - in the right level school. 1200 isn’t “poorly prepared” among typical graduates. It is in top schools. IF they managed to get in, they’ll likely still graduate, but with the massive pre-med culling that goes on few will head to med school often because they decide they can’t make the cut and quit trying. Put the same student in a different school and the outcome often differs. (Well, technically not the same student as one can’t do that IRL, but very similar students and seeing the outcomes.)
My D25 (yes, seriously) has been laser-focused on majoring in philosophy since she was seven years old. (It’s based in a deep desire to know not just how people behave the way they do, but why. I also blame Existential Comics.) Assuming that holds—a long shot, but hey, if it’s lasted five years so far, why not?—should I worry? No, because I also know that she has a pretty intense drive to succeed, and if she maintains that, then she’ll develop skills as a philosophy major that’ll be useful in a number of different career paths—and as a bonus, she’ll have a pretty firm grounding in what makes people tick. So even if she doesn’t do the common philosophy-to-JD path (doesn’t seem likely her thing, but it’s early days yet) or double major with something more “marketable”, market analysis consulting, anyone? Advertising? Lobbying? Policy analysis? Business or political communications? Mediation?
And that’s a few of the decent-paying possibilities for a philosophy major—and let’s be honest, if a philosophy major can swing a decent-paying job, it’s pretty much open to anyone.
Philosophy does not have much in major specific jobs, but philosophy majors may be more competitive than most for major agnostic college graduate jobs, since their major requires both humanities thinking and logical reasoning (many college students avoid or minimize courses/learning in one or the other).
Over the years I’ve worked with many programmers. Several had undergraduate degrees in philosophy. They were excellent. Very thoughtful and logical problem solvers and more importantly just nice people.
Perhaps a better way to group majors is not STEM vs non-STEM, but how heavy and how much math is embedded in the discipline. Math is hard, so the harder the math, the harder the discipline.
The law prof’s article boils down to the following:
-
If you believe you have world class math skills, go into math heavy disciplines. If not, then look at math light disciplines. That means humanities and to a lesser extent, social sciences.
-
If you want to be earn a good income without graduate school, take the most math heavy of the math light programs you can handle. He used accounting as an example. Mine went into marketing.
-
If you don’t mind doubling down and go the grad school route-law and to a lesser extent, business school- then you must adopt an entirely different strategy and aim for the highest GPA possible. So concentrate on cultural studies, gender studies and the like.
I find his reasoning clear, simple, and practical. I came to the same conclusion(s) a decade earlier myself. What is there not to like?
Canuck-If you believe that employment is divided into math vs. no math than the logic makes sense. The real world is MUCH more nuanced.
I was a humanities major who ended up in business school and had to take a lot of analytical/quant courses.
Guess what- humanities majors CAN out-perform their “mathy” classmates. Financial analysis- it uses 10th grade math. My brilliant, mathy, overprepared classmates often did not do as well as I did. Why? They were looking to use really complicated formulae to explain some pretty simple phenomena. I wasn’t “smart” enough to make up a complicated formula so I got to ask myself “Does this make sense” and usually that led to the right answer. Operations Management-- one of the “mathier” requirements in the B-school curriculum-- the better prepared students often did the worse. They wanted hard, complex, abstract problems to solve to show off their skills and training and smarts. Me? I just wanted to make the line shorter at the grocery store or eliminate the wait time at the cash machine or get the yogurt from the production line to the supermarket before it spoiled, so had to figure out the most efficient and cheapest and fastest way to do that. I wasn’t trying to win a Putnam.
I did not major in humanities to get the highest GPA possible (Greek and Latin are tough at the college level- and there are no shortcuts. If you are handed a passage from Herodotus you’ve never seen before, you either translate it correctly and poetically or not- and if not, you fail.) You sure have a dim view of what humanities teaches at the college level.
I remember how stunned my programming professor was during my first semester in business school. I was the only one in the class who had never used a computer (not once. This was “back in the day”, laptops weren’t invented yet, there was no excel or Word). He finally came to the conclusion that studying foreign languages in college was likely just as good- and sometimes better- than having actual programming experience. We’d get an assignment and my classmates all were trying to outsmart the computer. Me-- just deal with the punch cards in the right order, following the rules of the language we were using. Who ended up with the simplest- least likely to crash- least likely to fail program? Me- the dumb one. I didn’t need to be taught that every programming language had its own rules. If you’ve studied Ancient Greek you know it’s not Sumerian, and that sending a letter to someone in Aramaic who can’t decipher Hebraic script is not going to work.
I don’t know Philosophy is pretty hard. I took it in college.
So, FWIW, my D is taking a 400-series (upper division) math class at UMich this semester and there are roughly 40 students in the class with only 3 girls. So, are all the females taking the easy humanities classes. :lol:
The “massive pre-med culling” is often not so “massive” at the selective private colleges I am referring to, where the majority of the class graduates with an A average. For example, AAMC stats indicate that 202 students from Yale applied to med school last year. Yale has a graduating class size of 1313, so roughly 15% of the class. I realize some may be applying a 2nd time or have other special situations, but for simplicity I’ll say “roughly 15%.” In the corresponding Yale freshman survey, 15% reported plans to pursue a career in medicine. Many of the class reported being unsure of their future plans, so again it is not a precise value But if 15% report being interested in medicine, and ~15% apply to medicine, that’s not suggestive of a “massive pre-med culling” or that its only the kids who are in the top 25% by both SAT score and GPA who are getting in to med school. However, students who choose less selective colleges where they are in the top 25% by SAT score often have a completely different experience, with a far more massive portion of the class being “culled.” Colleges where more students get A’s, more students do A quality work, and few students do culling quality work tend to have far less culling.
I also have personal anecdotal experiences. I was a pre-med student at Stanford during a period in which Hum Bio was Stanford’s most popular major, and a large portion of the class was pre-med. I knew many pre-med students, took classes with them, discussed exam grades and plans, studied together, etc.
Most of my classmates persisted to medical school, including both the students who were relatively better academically during HS compared to typical Stanford students and relatively worse compared to typical Stanford students. For example, the pre-med student I was closest to was rejected to all the other colleges he applied except Stanford, including multiple UCs. One contributing factor for his Stanford acceptance is an advantage that is hook-like, but not a traditional hook… Given his many rejections to UCs and otehr schools, I expect he was towards the bottom of the Stanford class by stats, yet he still went to medical school.
The minority of pre-med students I knew who did not persist often pursued graduate degrees in other fields, and I have no reason to believe that they would have been rejected from med school had they applied. For example, one student I knew well went on to get her PhD and became a professor at an Ivy League college. Based on what I know about her, I think teaching was a much better fit than medicine.
I was in the bottom 25% of my Stanford class by both combined SAT score and overall HS GPA. However, I was in the top 25% of my class by both math/science SAT score and HS math/science rigor. This relates to what I said earlier about many factors besides combined SAT score contributing to success in college. SAT score is correlated for the typical student with a balanced score, GPA that his typical for his score, course rigor that is typical for his score, and all kinds of other key driving forces that are typical for his score. But if you don’t have those typical characteristics, then the expectation is often completely different.
During HS I had planned to pursue engineering and had no interest in medicine. However, I really liked freshman chem and did very well in the class, so it inspired me to continue with the pre-med track. I took the most rigorous sequence of pre-med intro math/science offered at Stanford, which is more rigorous level than the vast majority of Stanford pre-meds, even though I did not meet the required pre-reqs for one of the more rigorous classes (required “mastery” of AP class, when I hadn’t taken AP class or equivalent). This wasn’t a disaster, although the class in which I did not do the pre-reqs was quite challenging. I got to be a regular at the physics tutoring place and regularly worked through assignments with grad students. I probably spent more effort on that class than any other one at Stanford, and was probably more proud of the A on my final (or maybe it was midterm) than any other grade I received freshman year. Based on my intro math/science grades, I have no reason to believe I would have been rejected from med school. However, ultimately I decided to pursue grad school in engineering, rather than medicine.
A 1200 student may or may not be “poorly prepared” for college. It depends on far more than SAT score. Similarly a 1200 SAT score student may do really well at both a more selective college or a less selective college. As discussed, it depends on many factors besides score.
@Data10 I just looked up Stanford’s SAT scores on Collegeboard:
https://bigfuture.collegeboard.org/college-university-search/stanford-university
Roughly 1% have a 1200 SAT score (or less)
Not surprisingly, Yale is similar:
https://bigfuture.collegeboard.org/college-university-search/yale-university
You and I are talking about different students. I’m talking about average 1200 score students and you’re talking about the 1% or so who can make it in to tippy top schools. Perhaps that 1% could make it into med school since something attracted Stanford/Yale to them though I’d be curious to know if any in the bottom 5% headed that route. Otherwise I asterisked out those who attend Top 20 places because there essentially everyone is of a similar caliber. Bottom 25% is really meaningless at these schools when it practically means 1300+ for most. (Only 12 - 14% have < 1400.)
There’s not much more I can say if you don’t feel there are differences.
Readers can gauge their own thoughts. Mine are based upon 20 years of IRL experiences from one statistically average high school with an annual graduating class of 300+/-. It’s a data point for folks to consider. Nothing more.
@blossom A hypothesis is by definition a simplification of reality. Most humanity students do not do classics or analytical philosophy, which are among the most difficult of the majors. This graph, alluded to by the Law prof, shows how different majors stack up:
https://www.discovermagazine.com/the-sciences/classicists-are-smart
I made a study of majors because I think they are better proxies for ability than one’s alma mater. I thought so even before the Harvard lawsuit.
Canuck- these are your words:
“Perhaps a better way to group majors is not STEM vs non-STEM, but how heavy and how much math is embedded in the discipline. Math is hard, so the harder the math, the harder the discipline.”
There is zero math in Classics. So Classics is easy, QED? Since you claim that the harder the math, the harder the discipline?
I’n not sure what point you are trying to make, except that perhaps people who are good in math are generally thought of as smart by other people, and people who are not good in math are generally thought of as dumb by other people? Surely you’ve met someone who is in a profession because they love it, even though they “could have” done something harder, more lucrative, considered higher status than what they do?
I have no idea whether the leader of my religious congregation is good at math. But he’s a gifted leader/preacher/listener/counselor/community convener. I don’t know ANYTHING about his innate abilities in math or science based on his demonstrable track record in a difficult, highly stressful profession.
Diving the world into Stem vs. non Stem, or how much math vs. no math seems like an illogical way to parse the world given how many professions don’t fit into your model. But hey, what do I know. Not a math major.
@blossom I think you are being argumentative. Can you not see that I started the sentence with “perhaps”? That means it is a proposal, and the process is iterative. When better empirical evidence is available, the hypothesis may need to be modified.
I will say it one more time. No hypothesis is perfect. The best we can do in the social sciences is in standardized testing, yet R is only .5 or a little higher (in predicting job success).
If you don’t like the evidence put forth in the graph, how about this one?
https://qz.com/334926/your-college-major-is-a-pretty-good-indication-of-how-smart-you-are/
Your original comparison was 1200 SAT students in a class with 1500 SAT students. This specific situation is uncommon. If the class is full of 1500 SAT students, there are generally few 1200 SAT kids. Using ACT conversion, a college with ~25% of class at 1200 and ~25% at 1500 would have an 25/75 ACT range of ~24 to ~34. According to IPEDS, there are not a single college in the United States that meets this description. All colleges that report >= 25% of students with 1500 report <25% with 1200.
I assume the principle does not require such an extreme and rare score mismatch that would apply to only a minuscule portion of students, and likely applies to extremely few within your anecdotal sample. When I attended Stanford, the 25th/75th conversion was 25% in 1300s or lower and 25% in 1500s or higher. I scored a 1300, which put me in the bottom 25%… not as extreme as 1200 vs 1500, but still a significant difference.
I’ve been using extreme examples to make the condition more clear, as I expect you have. It’s easier to see the effect if you look at extremely selective private college where the majority of students graduate with an A average, but if you drop down a few tiers the similar principle applies. The portion of the class getting A’s is well correlated with the portion of class doing A quality work, and the portion of class that is “culled” is well correlated with the portion of the class doing culling quality work. More selective private colleges tend to have a smaller portion of the class “culled” than average and a larger portion persisting within pre-med and ultimately applying to medical school. They might not have 90% of the class graduating with low A or better average, as suggested by the Harvard senior survey. But in general, the portion receiving A’s and culling behavior is well correlated with selectivity. Of course there are many additional factors in pre-med attrition beyond grade distribution, and certainly beyond how combined SAT score compares with peers.
You are assuming everyone has a balanced SAT score. It’s quite common for scores to be unbalanced, in the lower range of an individual sub-score. For example, Stanford might be more forgiving of a lower CR if the student aces the math section, especially if that student is planning a STEM major. This distinction between subscores and composite score is important if you are using the score to guess at future performance in math/science classes. But yes, I agree that the vast majority of current Stanford students have >1400 scores. The actually figure for class of 2023 was 16% with 1200-1390, 1.3% with 1000-1190, and none below 1000.
“Perhaps” your hypothesis could be modified to include a large swath of the “very smart” folks in the world who major in, and professionally work in, fields which you consider to be lightweight/the province of the “studies” crowd.
I have no trouble accepting that most theoretical physicists are “smarter” than the typical early childhood educator or dental hygienist. I push back at the extension of your theory that someone who is a social worker is ipso facto dumber than a programmer or IT Manager. Not every programmer is Cal Tech or Stanford material- and I’ve met loads of dumb ones who “plug and chug” and couldn’t handle a day as a social worker because they lack the intellectual bandwidth to solve problems in real time that don’t always have a correct answer. And spend ten minutes with your typical corporate “IT Manager” whose parents are thrilled that he/she majored in STEM instead of a “loser” major like Asian Studies or Renaissance History, and that typical IT Manager will disabuse you of the notion that IT is a science requiring facility with the scientific method and numerical analysis. And will make Renaissance History start to look pretty good as a discipline which churns out well read graduates who can write, think, and figure stuff out.
So your comment that perhaps the world can be divided into “math heavy vs. math light” can get modified into “there are lots of very smart, intellectual people who in fact, do not study “math heavy” subjects in college. And that doesn’t detract from their analytical capability one iota”.
And then you and I can stop arguing.