A Limited Defense of Weed-Out Classes
STEM majors really are hierarchical
Via Matt “Dean Dad” Reed, I read this Washington Monthly piece on a study of wage gaps for different undergraduate majors. Specifically, it’s looking at a study from the University of Texas system that showed the usual large racial gap in pay within STEM and business majors, but not for non-STEM fields. The key figure is here, showing median salaries for different career stages broken down by race and major:
While there’s a significant and consistent racial gap in salaries for majors in the “computers, statistics, and mathematics” category and likewise for business majors, there’s little or no gap in the “humanities and liberal arts” category. The bulk of the piece is devoted to trying to find a plausible explanation for this difference in gaps that is flattering to the “humanities and liberal arts” fields. (As opposed to “There’s kind of a floor on what you can pay someone who has a college degree, and all three groups are running up against that.” Which seems pretty plausible to me, given the vertical axes on those plots…)
The factor they give the most weight to is the existence of “weed-out” courses in the STEM fields. The key paragraphs are here:
One of the key differences between humanities students and their counterparts in science, technology, and math shows up in the first year of college. That’s when many STEM, business, and pre-med students are herded through notorious “dropout” or “weeder” classes—big, required courses with extremely challenging curricula that are purposely designed to cull any students deemed less prepared to withstand the rigors of their chosen majors. “The failure rate, especially for minorities who come into the sciences, is pretty high,” says the physicist Walter Massey, president emeritus of the School of the Art Institute of Chicago and of Morehouse College, and the former head of the National Science Foundation. “Within the humanities, there’s not that sort of weeding out.”
By itself, that doesn’t explain the earnings gap, which measures the incomes of students who do make it through weeder courses and graduate with degrees in high-wage fields. But it turns out that the stresses of these courses can damage the career prospects of students who stick it out, especially if they arrived at college from underperforming high schools (as many students of color do). “In their first year, they need to make this really hard choice—to try to simultaneously catch up in calculus and take Computer Science 101 or take remedial calc, which delays graduation by an entire year,” says April Christina Curley, the engagement and partnerships manager at the Last Mile Education Fund. She says students who struggle with weeder classes may be denied access to high-level courses in specialized topics that are prized by leading employers, or they may self-select into less rigorous elective classes. Their initial difficulties also can undermine their self-esteem and dissuade them from applying for jobs with the most highly selective employers.
Later on, the key recommendation also concerns these courses:
Curley and others say one good first step might be to dump “weeder” classes, or at least create some workarounds so students who don’t make it through the gate the first time get a fair chance (including tutoring and other supports) to try again. These reforms should include retraining instructors to focus on finding ways to retain students, instead of just warning entering freshmen, “Look to your left. Look to your right. One of you won’t be here next semester.”
And, look, I absolutely agree that the “Look to your left, look to your right” shtick is gratuitously dickish. There’s absolutely no reason for that kind of hostility, even in a large-lecture-hall sort of scenario. As a profession, we should absolutely be making efforts to be more welcoming and supportive to introductory students.
At the same time, though, saying “Just dump ‘weeder’ classes” always gets my back up a little. It often seems to come from a belief that the “weed-out” courses early in STEM majors1 are something we do just to be dickish and exclusionary, a belief that’s both widespread and remarkably resilient. I’ve spent a lot of time over the last twenty years trying and failing to get colleagues across campus to understand that science and engineering majors really are hierarchical by nature, in a way that doesn’t readily lend itself to re-ordering courses on the fly. I’ve had virtually the same conversation with the same people multiple times with no discernible impact on their views the next time the subject comes around; as an incurable optimist, I will now write about the topic on the Internet, because I’m sure that will fix everything.
When I say that STEM majors are inherently hierarchical, this is not to say that the existing curricular structure is graven-in-stone TRVTH that can’t be changed— it can be useful and illuminating to reconsider the ordering of topics in the intro sequences, or to offer intro sections with different “flavors” that present the key concepts in different contexts. We used to use the “Matter and Interactions” curriculum for intro physics, which presents Newtonian mechanics in a very different way, and I liked that a lot (this opinion was not widely shared, so we’ve changed back to a more traditional format). And our CS department does several different versions of the first course in their major sequence— a version framed around “Big Data,” a version framed around game design, etc.— which I think has worked very well for them.
While there can be, and often is, significant disagreement about what order to cover things in, and which skills are most fundamental to later success, there will necessarily be some hierarchy to STEM majors, because they require conceptual and technical foundations that can’t readily be picked up in high school or by independent study. You really, truly can not walk into a 200-level course in most of these majors without having picked up the material that’s in the 100-level courses. And not every student will find those concepts and techniques congenial in a way that will lend itself to success in those majors, no matter what order you put them in, or how gently you introduce them. If you want to be a physicist, you’re going to need to learn vector calculus at some point, and vector calculus is not a subject that’s equally appealing to everyone.
It should also be noted that this process of “weeding out” the students who don’t find the foundational elements of a discipline congenial is not in any way exclusive to the STEM fields. I’ve written before about my freshman-year experience with a 200-level English course— an “Intro to Medieval Literature” course that was also a survey of various critical theories— that very effectively weeded me out of that field. This was partly because the professor was gratuitously nasty in her comments on papers (I got “This might’ve been a good first draft of a paper had you put any thought into it at all”), but also because I realized that the mode of thinking required to do well was not something I enjoyed.2
This is not, I hasten to add, intended to disparage the literary disciplines, just a recognition that they’re not to my taste. I had somewhat similar reactions to intro courses in Philosophy and Religion and Political Science, though those professors were much more pleasant to deal with. Those are all worthy subjects, and the people who study them for a living are doing real and meaningful work; it’s just that the things they do in the course of that work are not my bag. Which is fine; I’m pretty sure they’re not huge fans of vector calculus, or even linear algebra.
So “weeding out” happens everywhere: in every intro course in every field, there will be students who think they like the field based on the version they saw in high school, but will find that they don’t enjoy the college (and higher) version. The problem is just more acute in the STEM fields, for two reasons. One is the hierarchical nature of the fields, but that’s really just an aggravating factor atop the main driver, which is the vertical axes on those graphs above. Salaries in the STEM fields are a factor of two or more above those outside STEM, which leads to a lot of students coming into intro STEM courses with dollar signs in their eyes who are not, in fact, people who will enjoy doing what’s involved in those fields. I deal with a lot of struggling students who say things like “My parents really want me to major in engineering, but…” I doubt that conversation is quite so common in, say, Art History.
This does not in any way excuse the actual hostility expressed in some versions of intro STEM courses, or the very real barriers put in the way of too many students from under-represented backgrounds. We absolutely should be doing more to ensure that people are “weeded out” only for valid reasons—that they genuinely don’t find the core skills congenial— and not for reasons of race or gender or other personal characteristics. We should work to make the kinds of personal connections between students and faculty who can act as mentors (recommended elsewhere in the article), to help avoid the kind of “hidden curriculum” problems they mention, of students not knowing to take certain courses to maximize their future opportunities.
(There are some resource-allocation implications of this that won’t be popular with some sectors of the faculty, of course…)
At the same time, though, just saying “get rid of weed-out courses” is too glib. We can and absolutely should get rid of the attitude that students in the intro courses are faceless masses to be winnowed down to find the truly worthy, but the core phenomenon of significant numbers of students finding that they’re not well suited to what they thought would be their major is always going to be there. The problem of needing to pass (and to some degree enjoy) course A before you can succeed in course B, then C, then D is inherent to the disciplines and not something that can readily be worked around.
So, yeah, I’m sure there’s nothing that can go wrong with writing this and sending it out into the world. If you find this interesting enough to want more, or want to foist it off on someone else, here are some buttons you might consider clicking:
If you just want to yell at me about stuff, well, there’s a comments section for that.
I’m going to focus on the STEM fields, because that’s what I know, and what I react to. They also find problematic gaps in Business majors, but I have no experience with them (other than knowing that when the School of Management guys showed up to play pick-up hoops at Yale you were more or less guaranteed to have nasty and stupid arguments).
I ended up with a decent grade in that class, because with effort I could turn out a decent pastiche of the kinds of things she assigned us to read, but it was not a pleasant process.
The thing about "weed-out" classes is that they're in some respects indiscriminate--to extend the metaphor, they sometimes spray so much curricular pesticide that they wipe out many people who might either *eventually* find the core skills and methods congenial or they underimagine what the core skills being laid out actually are in relationship to all the professional activities that have some relationship to that core.
The former point has become more and more intense for me over the years--as STEM departments have been under pressure to improve outcomes for underprepared students and to make their majors more inclusive, there have been more students who survive a "weed out" despite receiving poor grades in the lower rungs of the curricular hierarchy and maybe only barely making it through the major at all only to flourish spectacularly at a later date in a professional or scholarly role that rests on their undergraduate education. It means there really shouldn't be courses that are impassable barriers, especially in terms of what you're suggesting here--that since you can't have students engaged in the actual practice of a STEM field before they have the necessary preparation, you have to give them some feel for how challenging that practice is going to be, whereas many of the humanities and social sciences can set students to actual practice of their fields and disciplines from the get-go and just help them become more and more sophisticated in that practice over time.
But I also think many STEM faculty under-imagine what those end-state professional practices are in relationship to the discipline that they themselves work with. The relationship between primary care medicine and biology/biochemistry is more distant than a lot of faculty credit; the notion that organic chemistry is giving premed students a challenging window into their professional future and letting them decide whether they're going to find that life congenial just seems empirically untrue. If I really wanted to "weed out" future doctors before they get themselves deep in debt and find it's not the life they wanted, I'd compel them to endure the terrible conditions facing many residents, I'd have them deal with insurance adjusters telling them what they can and cannot offer as a standard of care, I'd have them spend a day listening to patients of all temperaments and presentations, and I'd have them spend a Saturday night in a crowded underfunded urban emergency room. The notion that organic chemistry is a good proxy test of what life as a basic physician (or other health-care professional) will be like just seems wrong. I think we try to tell ourselves that our assessments are good predictions of later professional competency, and I think for our own disciplines, that's absolutely true. I cannot imagine that someone who profoundly hated physics (or history) could perservere and become a skilled scholarly physicist or historian. But I can imagine someone surviving biology enough to do any number of jobs where biological science is somewhere present but not determinative of everyday professional practice. In some cases, this is not at all hypothetical--there's tons of evidence that barely-passing CS majors may end up being enormously successful tech designers or executives, barely-passing engineers may end up flourishing in industry or policy, etc.
I agree with you that STEM fields do build on foundations, which end up serving as weed out classes. That being said, I also think the way of teaching physics that I experienced was designed for future physicists and not for a broader audience (eg. Engineers, other majors taking intro Physics). They often dove into the math rather than taking the pains to explain how it connects to real life.
Moving further into the course, us experimentalists experienced something similar from courses taught by theoreticians. For instance, we were taught Jackson, without explaining why and when the formal framework is needed. I finally got "weeded out" when I went for a QFT course in grad school, and even the first lecture wasn't explained in a way I could understand.
In professional life, I have found people of all backgrounds pick up a working knowledge of almost anything that relates to what they do. I think physics education has a lot of opportunity to make it more accessible to a wider audience.