More try STEM majors — and quit

More students are trying — and quitting — STEM (science, technology, engineering and math) majors, reports USA Today.

Interest is up, says UCLA Professor Mitchell Chang. Persistence is not.

Many students aren’t prepared for the rigors of introductory chemistry and calculus, says Clemencia Cosentino de Cohen, a senior researcher at Mathematica Policy Research. Women are more likely to drop the major.

“If women get a B, they think they’re failing. A man gets a B, and he’s happy. They say they’re acing the class,” Cosentino says. “Women who go into hard sciences, they’re very driven, they’re very high achieving, and if they’re not performing at that very top level, they become discouraged, and they think that it is not for them.”

Tough grading in science classes leads to attrition, a 2010 Cornell study found. STEM students realize they can work less and earn higher grades in liberal arts courses.

The S in STEM has been oversold, writes Washington Post columnist Jay Mathews.

“Employers are paying more, often far more, for degrees in the fields of technology, engineering and mathematics (TEM),” College Measures President Mark Schneider wrote in his report, “Higher Education Pays: But a Lot More for Some Graduates Than for Others.”

But “evidence does not suggest that graduates with degrees in biology earn a wage premium — in fact, they often earn less than English majors,” Schneider wrote. “Graduates with degrees in chemistry earn somewhat more than biology majors, but they do not command the wage premium typically sought by those who major in engineering, computer/information science, or mathematics.”

A TEM bachelor’s degree qualifies a graduate for a good job. An S bachelor’s degree usually isn’t enough on its own, though it can be the first step to a medical degree.

About Joanne


  1. Whether we like it or not, the future of our country (in fact, the future of any country in the 21st Century) depends greatly on our STEM skills as a nation. The nations that fail at STEM will not be world powers by the end of the century, no matter how big physically they are, or how many people they have. (The USA is #3 in size [only Russia and Canada are bigger… and maybe China, if you let them count Taiwan] and the USA is #3 in population [only China and India have more people].)

    If we want to remain a global power, much less the world’s lone superpower, we’d better find a way to make STEM happen on a much bigger scale among our population – and really happen, not just fake ‘grade inflation you really tried so here’s your piece of paper saying you tried’ happen…

  2. So STEAM would be better as TEAM? That works.

  3. Freshman weeder courses used to be the norm, on a university-wide basis, not just in STEM. At my flagship state university, they they were the freshman sciences, freshman English (lit-based comp) and precalc/calc. The aim was to weed out 1/3 of the freshmen by the end of the year and it varied drastically by intended major. In engineering, med tech and nursing, only about a third survived in their majors, and most transferred into easier majors (often into education, in my experience and that of my brother). This is nothing new. In fact, weeding out kids unlikely to succeed is doing them a favor, since they can transfer to another major easily as a freshman and far less easily as a junior. It’s also nothing new that women are likely to make different choices than men, and government intervention is not needed. A niece just received her doctorate in biomed engineering this year and is beginning her first job at the age of 29; not something that appeals to most women. Even so, if she wanted to go into academia, she would have had to take a post-doc fellowship, further delaying entry into full-time employment.

    • I think that under Obama’s DOE rules, weed out courses are illegal… If not de jure, then de facto…

      • Probably; disparate impact and all that. I’m pretty sure that mismatch is at least part of the problem. I’ve read a study done at Duke on the attrition from STEM, and minorities are much more likely to transfer out. If admitted with weaker backgrounds, that would be only logical.

  4. I remember my first computer science class in fall of 1981, it was CS 116 (Fortran) and we averaged 2 to 3 homework assignments a week, a quiz every two weeks, and we had 4 exams and a final project (write the parser for a airline reservation system, which we got handed to us when some idiot student complained to the department chair that the homework we were being given wasn’t relevant).

    On top of this, we had to learn a whole new mainframe (Operating System, editor, and compiler/linker), and after
    about 4 to 5 weeks, if you weren’t figuring it out, you were
    filling out a drop slip and giving serious consideration to
    changing majors.

    Weed out courses are designed to separate the men from the boys and the women from the girls…

    The reason why the attrition rate is so high in STEM courses/degrees is that by definition it’s HARD, not easy (contrast the discussion a while back about education majors).

    Go Figure

    • Bill: My recollections (and those of friends) about studying physics, math and the like jibe with yours.

      But then I keep seeing surveys about how much kids study in college …

      Now, the difference between engineering and education is maybe 4 hours per week.

      How do we make sense of this? Assuming OUR recollections are good and the these anecdata are in fact valid data, either we’ve got ed majors over reporting, engineers under reporting; or perhaps a qualitative difference in the type of student?

      I (tentatively) accept as a given that engineering and hard science homework is more intellectually demanding than playing with sock puppets (an example given in the ed major is easy thread); but we could examine that if you wish.

      • Mark Roulo says:

        What leaps out at me is not the difference between majors, but the absolute numbers: “The average full-time undergraduate student studies about as much as faculty members expect—15 hours a week—but the duration varies by major, according to this year’s National Survey of Student Engagement, released today.

        Engineering majors spend the most time studying, 19 hours a week…”


        Figuring 15 units as a typical load, so 15 hours/week in class … this means that the typical college student is putting in 15+15=30 hours/week on academics.


        This seems low. I remember a rule of thumb of 2 hours of study for every 1 hour of class, so 15 units would be 45 hours per week instead of 30.


        My local public high school runs 30-35 hours/week just of class time (maybe some of the classes are “study period”?). With 2 hours/night of homework, the high school students are putting in more hours per week on academics than the average college student.


        Maybe there is a large spread here? 40+ hours/week at places like Cal Tech and fewer at many other colleges?

        • Mark; I’d say that the places like Caltech aren’t much more than a rounding error in the averages. Their numbers of students are totally dwarfed by “the others”.

          Should you get the raw data to analyze, be aware you’ll have to figure out how to deal with the fact that Caltech was (is) under NCAA sanctions for academic violations.

          – so students would have more time to study since they aren’t practicing as much.

          The classic Carnegie Unit (foundation for high school “credit”) is for 120 seat hours – let’s say 120 lessons and about 15 minutes of homework a night (30 hours). Sooo, 150 hours per course over a year. College course – 3 class hours a week, 13 weeks? 39 hours plus 2 hours per week outside effort per contact hour (in some world) – 78 hours. 117 hours for the semester course. College is “flipped” – kids are taking more responsibility. Also, there’s likely to be some selection pressure – those more inclined (able?) to do seat academic learning?

          Now, I remember some labs that were 10 hours a week and only 4 credits. And the write ups – UGH. Yeah, recreate the Stern-Gerlach experiment; starting with a magnet, AC current and a pack of chewing gum. Gee, thanks.

          • Mark Roulo says:

            Sean: ” I’d say that the places like Caltech aren’t much more than a rounding error in the averages…”


            Well, Cal Tech is, but then I’d add MIT and Georgia Tech and UIUC. And then I’d start considering tough majors from lots of other schools (e.g. UMich engineering … 15 hours/week of homework … really? Same question for Princeton physics … or maybe physics at *any* “real” school). Then we’ve got reading/writing heavy majors at good schools (e.g. English literature at Yale?). I’d expect more than 15 hours/week if one was a literature major at a good liberal arts college.


            I can also imagine some schools with almost no homework at all (e.g. a “general studies” major at Directional State U. or wherever). So maybe we can get a 19 with a lot of 10s (at pretty bad schools where most of the kids probably shouldn’t be in college at all) and a small number of 30+s?


            NOTE: Thanks for the Cal Tech link. I was aware of it … and think that Cal Tech got off lightly! No wins vacated? Come on …

        • Deirdre Mundy says:

          I went to a demanding college, and was amazed at how much less time I spent on school than at my Magnet HS. Of course, my ‘commute’ went from an hour each way to 5 minutes each way, and I went from 8 hours a day in class to more like…12 hours a week. So even though I had more homework, I had more time to do it in, too…

          And the homework involved fewer projects, more problem sets, papers, and reading assignments, so it was easier because there was never any cutting or pasting or hand lettering involved…….

  5. The foundation for success in STEM fields, and in other demanding fields with tough entry standards and tough internal standards, is laid – or not – in k-12. Mostly, it’s not. The ES curriculum (and often the MS) appears to be an epic fail (with some exceptions, probably highly correlated with parent SES and availability of outside help) at the task of establishing a strong foundation for deeper learning in the HS years. It should also be noted that the CC does not provide for STEM etc. readiness.

    • I’m betting mostly on ES, by middle school the “clay” is starting to get a bit firm.

      I’ll go with the availability of outside help – tutoring places around where I live do a brisk business. (Now, that costs money, but what’s more important is the desire to spend it on your offspring and their needs.) Couldn’t believe the throng of parents hanging outside Kumon Centers on Friday night. Almost like the throng waiting to eat at Outback Steakhouse. Well, in the thronginess at least.

      SES certainly comes into it, anybody have links to some good studies? I’d have thought expensive prep schools would be a nice proxy; but often their scores and prep don’t seem so awesome for the expense and selection involved (hot house effect) when compared to affluent suburban districts.

      • “High-performing” schools deliberately avoid considering the “outside help” issue (whether parents, Kumon, paid tutors etc); they do not want to know. They point to measures of student success and claim it’s because of their school. The fact that my second-grade grandtwins, in a high-performing and highly competitive public school, have their nightly homework supervised and checked by their parents, with assistance provided as needed, is never acknowledged. Their teachers do ask the kids to read aloud to an adult for 15″ a night, but the homework (math, spelling and writing) is designed to be done by the kids alone. However, parents and tutors do check and correct and my son and DIL provide lots of extra challenge work. In fact, that is the norm for their school. Due to parent pressure, the school ditched Everyday Math, in favor of Singapore Math, several years ago – because of the flaws in Everyday Math and the absolute necessity of full-time tutoring. The local Kumon director, in an article about the curriculum change, said he’d have to make some changes in his marketing.

        • momof4: Confirms my experience. Last year I had a 2nd grader. She had a dreadful math curriculum, it might have been Everyday even. We supplemented with Saxon 3 at home. Then they started this Words Their Way spelling program – instructions home indicated that adults were NOT supposed to help. Kids had like 20 words a week. Each night they were supposed to do an exercise – like letter bingo, or finding words with similar patterns or arranging words in pattern groups; typically with just 5 of the words on the list. Class performanace sank like a rock for a couple weeks – until parents took over and went back to drilling spelling words.

          The school declared success and moved right along.

          It’s been a while since Joanne had anything on the board about how many STEM people wouldn’t want their kids to major in it; or at least work in the field. That to me is discouraging – people already into it not wanting to foist it off on their kids. In my mind it indicates a pretty deep problem.

  6. Of course more students are interested but can’t hack it. Science classes in K-12 are becoming edutainment where kids do cool hands on stuff but don’t learn any science behind it. Students always tell me they loved science in middle school. I ask why and they tell me something cool (and I agree) that they did in class. I then ask what they learned and I get a blank stare.

    Students have become convinced that scientist sit around blowing things up. Those hours spent working through difficult problems, writing and writing, and reading are basically hidden from the students to ensure they might considering majoring in STEM.

  7. In my experience, the barrier is math. If a major requires lots of math, then there will be a steady stream of drops and changed majors. My aerospace engineering class started with about 200 folks in “Intro to Aero 101” and graduated 16 folks four years later. It wasn’t that we 16 were smarter than the rest, not at all. It was that we were the ones willing to do the work (perhaps we were the dumb ones). The others realized that they could change to easier majors and did it.

    Most engineering courses, like it or not, are problems-based courses. The thing that you K-12 teachers like to call “drill and kill”. There simply isn’t any substitute for actually doing the problems and getting them right.

    For example, courses like “engineering statics” or “engineering dynamics” (common to nearly all engineering curricula), the standard was 10-20 problems, each taking 10-15 minutes to figure out, due each class period, all semester long. It’s a crapload of work, but you certainly learn something.