Why Science and Engineering Majors Change Their Minds

Monday, November 7th, 2011

It’s hard not to get a chuckle out of the New York Times reporting that science and engineering majors change majors because their classes are so hard:

The excitement quickly fades as students brush up against the reality of what David E. Goldberg, an emeritus engineering professor, calls “the math-science death march.” Freshmen in college wade through a blizzard of calculus, physics and chemistry in lecture halls with hundreds of other students. And then many wash out.

Studies have found that roughly 40 percent of students planning engineering and science majors end up switching to other subjects or failing to get any degree. That increases to as much as 60 percent when pre-medical students, who typically have the strongest SAT scores and high school science preparation, are included, according to new data from the University of California at Los Angeles. That is twice the combined attrition rate of all other majors.
The latest research also suggests that there could be more subtle problems at work, like the proliferation of grade inflation in the humanities and social sciences, which provides another incentive for students to leave STEM majors. It is no surprise that grades are lower in math and science, where the answers are clear-cut and there are no bonus points for flair. Professors also say they are strict because science and engineering courses build on one another, and a student who fails to absorb the key lessons in one class will flounder in the next.

After studying nearly a decade of transcripts at one college, Kevin Rask, a professor at Wake Forest University, concluded last year that the grades in the introductory math and science classes were among the lowest on campus. The chemistry department gave the lowest grades over all, averaging 2.78 out of 4, followed by mathematics at 2.90. Education, language and English courses had the highest averages, ranging from 3.33 to 3.36.

I’m not sure I agree with this commenter, but I enjoyed how he made his point:

People who study science want to become scientists. People who study engineering want to become engineers. People who study liberal arts want to get a degree. This is the difference.

This re-raises the question of what all these degrees and GPA-measurements are for.

If you already have a technical degree, you probably don’t want to see its value degraded by a new “dumbed down” curriculum. (Fortunately, many schools already offer an “information technology” degree to avoid diluting their computer science degree, but I don’t know of any “chemical engineering concepts” degrees.)

On the other hand, if you do technical work, you probably don’t use even the tiniest fraction of what you learned in school. Very little of that rigorous and rather theoretical mathematical analysis makes itself useful.  And the key concepts you learned? Well, all those B and C students, who got degrees just like yours?  They didn’t learn those key concepts.

The same classes that test a student’s grit and intelligence may not be the ones that teach useful skills for actually engineering something.


  1. David Foster says:

    If the alternative to a STEM degree was a major that required learning Latin and ancient Greek, or that required the close reading and analysis of medieval texts, then things would be a little more balanced.

  2. Ben says:

    Making broadbrush polarizations (“liberal arts vs. science”) seems fun, but David Foster’s comment is more to the point.

    Let’s see if we can say this in a “Tao” way. “It’s not about the discipline (course work), it’s about the Discipline (level of effort applied.)”

    Try reading some early/mid 20th century Classical Arabic inguistics in the original German; or get an MFA in oil painting from a respected school. It can be real work.

    As for skill re-use: one may not do the exact same exercise later — or ever again — but learning how to do it the first time, and how to do it well, later, confers skills that last a lifetime.

  3. Isegoria says:

    I completely agree with David Foster’s point that a non-technical degree could be made extremely challenging — but, if you’re going to encourage more and more students to go to college, and you need to lower standards, then the less technical fields without clear right or wrong answers will naturally grow to fill that need, and grow less and less objective and rigorous in the process.

    A century (or more) ago, mastering Latin and Greek meant that you had the intelligence and grit to do great things, and you might thus earn a position running one of Britain’s colonies — even though Britain obviously did not rule over Ancient Romans and Greeks.

    Today’s technical degrees and advanced professional degrees work much the same way — not just for future politicians and business leaders, but even for technical workers. You don’t learn how to engineer at a university. You demonstrate your ability to knuckle down and learn some difficult math associated with academic engineering. And you don’t learn how to learn at a university, either. Researchers desperately looking for that effect have been unable to find it.

  4. Jehu says:

    What a person in one of the more mathematical disciplines of engineering (e.g. Electrical or Mechanical) learns in their post-calculus engineering mathematics courses is this:

    In the first course post-calculus (often called Engineering Math I or Ordinary Differential Equations) is how to do ordinary arithmetic and algebra really really well. Those basic foundational tools are exercised over and over again nearly invisibly when working problems in Engineering Mathematics. The same is true of calculus in Engineering math II (or Partial Differential Equations). This makes math up to the level of basic calculus available to the gut of a good engineer. It makes them truly numerate. There’s a difference between knowing how to use calculus and having it so ingrained that it has become available to your gut.

  5. Ross says:

    “…don’t learn how to learn at a university, either…” Mmmm… don’t know about that, Isegoria. Maybe most don’t? I know I did, but the real learning about learning didn’t start until very late in undergraduate, and I hit my stride in grad school. Definitely learned skills on how to read, think, study, take notes, discuss, research, drill into citations, etc.

    It’s often dangerous applying generalities to aggregates. Perhaps it’s true most don’t learn to learn at school. I know, however, that’s where I started.

    As for Jehu’s observation: 100% Spot On. I think that may go to the “10,000 hours of practice” meme.

  6. Red says:

    The other problem with STEM is the resulting job (beyond a teaching gig) is a low-status job. Chicks do not get ga-ga for Engineers and geeks. Money does not equal status is our society.

  7. Matthew Walker says:

    Red, it’s all attitude and social skills. STEM jobs happen to correlate well with wrong attitude and crap social skills. If you turn a woman on, she’ll find some way to explain away any reservations she has — about your job, your felony convictions, your hump, or anything else. I’ve known programmers who banged some high-quality p—-.

    Not me, though. It’s the social skills.

  8. Isegoria says:

    I completely agree that mastering a subject is fundamentally different from just getting the gist — but you’ll note that a technical degree does not require mastery of the technical material; it requires a C average. I don’t know of any program that requires a 95-percent score on the Calculus 101 final to move on to the next class.

    But, to return to my main point, there is an important distinction between demonstrating your smarts and grit by excelling in those difficult classes and learning the most important skills for a technical career. Real engineers do not spend 95 percent of their time doing math, but engineering students do.

  9. Massive government subsidization of education has warped the transmission of craft from experienced practitioners to inexperienced learners.

    Today’s employers expect someone else, anyone else, to pay to train their employees, either by direct government appropriation to education or workers paying for their own education (meaning indirect government appropriations to education). In olden times, employer had a direct tit for tat relationship with employee where the employer provided training in exchange for labor. This employer provided education took many forms, from the knout to apprenticeship, but it was direct.

    One brother has the impression that the current unemployment rate was mostly caused by lazy people living on unemployment insurance since his large DJIA listed company had a still unfilled position after nine months. I asked him if they’d considered training someone, perhaps a junior employee, for the job. My brother said they were considering it training one junior employee but management was balking at the novelty.

    Most of the craft of software engineering I learned outside my degree path. Doing lower electives at the local community college helped since some of the adjunct professors were working software engineers but instruction by tenured faculty was often useless. Many of the software engineers churned out by the program came out with no skills other than the ability to provide textbook correct answers in bad Java. I’ve been in many workplaces where the craft has not been passed since the young kids don’t even know there is a craft.

  10. Isegoria says:

    Today’s employers have little expectation of reaping the rewards of training their own employees, because they don’t have any long-term agreement or expectation that the employees will stay.

  11. It’s an interactive process: employer shows a smidgen less loyalty to employee, employee shows a smidgen less loyalty to employer. Employee shows a smidgen less loyalty to employer, employer shows a little less loyalty to employee. And so round and round it goes. Where it stops is probably inland China, with both employer and employee replaced by their cheaper Chinese equivalent.

  12. Alrenous says:

    The problem I’ve always had with this is that you can demonstrate just as much grit learning useful things. Veblen suggests, however, that being able to afford to learn useless things is higher status.

    But imagine what you could do if you could afford a useless Ph.D. but put those resources toward something useful instead.

  13. Bruce Charlton says:

    Note: Employers never paid for the training of employees. Under the classical system of apprenticeship the apprentice paid the Master a fee to be taken on, and worked for nothing — often for seven years.

    However, once he had been allowed to join the Guild, the apprentice (now Master) had the benefits of a protected trade, with fixed prices, control of competition, etc. A guaranteed livelihood — so long as you were prepared to work and obeyed the Guild.

    Indeed, much like medicine still is.

  14. David Foster says:

    Perhaps it’s reasonable for an employer to have to train someone how to use a micrometer; it’s not reasonable for the employer to have to train him how to add & subtract decimals and fractions so that he can actually measure something with it. It’s reasonable for an employer to train someone in the latest version of a programming language; it’s not reasonable for the employer to have to train him in the conceptual basics of programming.

  15. David Foster says:

    By the way, GE is trying out a manufacturing apprentice program in the jet engine biz.

  16. Alrenous says:

    Not the first time GE has figured out the head-slappingly obvious management improvements.

  17. Isegoria says:

    I wouldn’t call eliminating all management a “head-slappingly obvious management improvement” — even if it’s the kind of thing every low-level employee dreams of. I’ve mentioned that Engines of Democracy piece before, by the way, as well as Joel Spolsky’s piece, How I Learned to Love Middle Managers.

    There seems to be a natural tension between top-down, centralized decision-making and bottom-up, decentralized decision-making, and middle-managers are, of course, caught in the middle. Napoleon made sure to have a group of trusted aides he could use as his directed telescope, to zoom in on important details by bypassing the staff hierarchy:

    As with every commander, Napoleon relied heavily on the routine reports of his commanders and staff. Nevertheless, he often found them lacking in detail, thus not fully useful. Also, unit reports frequently neglected to express the more intangible aspects of a unit’s status — like leadership, morale, and esprit. If subordinate commands did address these vital areas in reports, they were often less than objective and sometimes totally biased. To keep him informed on the significant conditions within his various commands and to verify his subordinates’ reports, Napoleon used his aides as a directed telescope to augment the regular reporting system. Napoleon viewed the formal staff system as totally inadequate when reporting what Van Creveld refers to as “less structured information.” The emperor-general also turned his telescope toward the enemy on occasion and on the terrain where campaigns were to be fought. It was Napoleon’s use of aides and his system of formal reporting procedures that contributed to his leadership techniques becoming a wholesale revolution in military command.

  18. George DeMarse says:

    We sure have gone a long way in this blog from how hard the engineering curriculum is to GE wanting to rid itself of management and calling this “headslappingly obvious.”

    Don’t be silly. GE, along with every other corporation, loves managers. They can justify higher salaries for managers and “blame” them for delivering bad news to the workforce–like downsizing and offshoring.

    The uses for managers by corporations are many and they are gold.

    George DeMarse
    The Sage of Wake Forest

Leave a Reply