[Part 2 of the Ed School Trials and Trivialities series, beginning here.]
As an undergraduate, I transferred late (by the pure chance of convenient scheduling) into what became one of my favorite courses, a general education seminar on the history of cosmology from the ancient Greeks to modern astrophysicists. The professor, an astronomer who specialized in galaxy formation, was a bit of a sour oddball, the opposite of the sort of cheerful, self-consciously nerdy types who like to appear on public-broadcast science programs. No, he told us sophomores eager to learn (to paraphrase a distant memory): every theory we cover in this course is wrong, Aristotle as much as Einstein.
I think he wanted to act as a provocateur – for example, he spoke highly of the accuracy of epicycles, the much-maligned medieval method for calculating the movement of planetary bodies, before demolishing them for lack of parsimony when we reached Kepler – so I don’t know how much he believed his classroom persona. But pedantically, he had a point: as a committed empiricist who rated theories less on their strict truth than their predictive power, he embraced a scientific ethic similar to that expressed in the statistical aphorism “all models are wrong, some are useful.”
Well, and some are just useless.
Take the psychological theory of learning styles, the idea that certain people learn better through identifiable mediums of instruction (often reduced to three or four: visual, auditory, read/writing, and kinesthetic). The theory predicts that matching a student of a certain learning style to instruction in that style will improve their educational outcomes. For this theory to have practical use, it must first overcome several broad hurdles to demonstrate itself viable:
1) Is it possible to identify a person’s learning style?
2) If so, does matching a person of a certain style to instruction in that style improve educational outcomes (termed the “meshing” hypothesis)?
3) If so, do the outcomes of matched instruction exceed the outcomes of unmatched or mixed instruction?
The first point simply asks if learning styles exist as an observable phenomenon. Or, in less precise words, is it even real? This may seem like a too-obvious place to start, but in the face of the continued popularity of scientifically dubious type theories like the Myers-Briggs Type Inventory, it is important to ask. Fortunately for learning styles, in this case the theory holds: when asked, students will reliably express a preference for certain styles of instruction and people possess measurable differences in aptitude of specific cognitive abilities like “verbal comprehension” or “spatial visualization” (see this literature review).
On the second point though, the theory begins to falter. In addition to the negative results found in the literature review already cited, further studies, including some with rigorous experimental designs, routinely fail to demonstrate significant differences in comprehension when meshing learning preferences with instructional practice. I don’t mean for this post to become a literature review itself but, in short, the empirical evidence for learning styles is contested and the predictive value of the meshing hypothesis is weak.
And then we hit point three, the practical test. Even if the theory surmounts the first two hurdles, is the finding useful? It is a question of application: are the instructional practices implied by the theory more effective than alternatives? Here, I see no reason to believe so at all.
In the interest of charity, consider a defense by this foundational scholar in the field who supports both type theories and the concept of learning styles. On the categorization of styles, he writes: “The fact that students may be classified as, say, sensing learners, says nothing about either their intuitive skills or their sensing skills.” But this is an absurd statement that obliterates his own theory. If a learning classification “says nothing about” a learner’s skills, why bother classifying them at all? A few sentences down, he also writes that “The optimal teaching style strikes a balance (not necessarily an equal one) between the poles of each dimension of the chosen learning styles model.” Aaannnddd, to which I ask again: if the recommended style for a mixed class of learners is a mixed balance of instructional practice regardless, why bother classifying them at all? It becomes a trivial theory.
To return to the wisdom of my astrophysics professor, I can give a false theory a provisional pass if it has strong predictive value, like epicycles mapping the movement of the planets. However, the trouble with learning styles starts before even considering the empirical evidence: one of its strongest proponents asks that it not be used for predictive purposes! By contrast, consider the competing theory of “dual coding,” which applies concept behind learning styles in a better generalizable, though incompatible, way:
… the two make contradictory predictions. … Learning styles would predict that visual learners learn best when they encounter visual information and auditory learners learn best when they encounter linguistic information. Dual coding, in contrast, predicts that all learners, regardless of their so-called learning style, retain information better if linguistic information is supplemented with visual information (source).
In effect, “dual coding” takes the conclusion of mixed instruction from the learning styles proponent above but removes the potentially costly baggage of measuring styles for little apparent purpose. Is the theory of learning styles true? I don’t know and, at this point, I don’t care. But I do know one thing: learning styles will have very little relevance to my instructional practice as a future educator. Against less ambitious but better applicable alternatives like dual coding, the theory is not useful.
And that brings me back to Ed School.
All of my professors love to talk about learning styles. They introduce themselves with their supposed styles (“I am a visual learner, so blah blah blah”), they ask us students to reflect on our own styles (I reflected “I don’t know.”), and, when discussing the otherwise sound concept of differentiation, they exhort us to consider the plight of that poor kinesthetic learner who just has to get up and move!
It gives my entire teacher training an unserious air.
I recognize that controversies exist within academia and, as I stated before, I have little interest in the strict truth of learning styles. But to push the idea so earnestly and not even mention the bountiful criticisms of an idea so often termed an outright myth! If they dismiss the evidence against learning theories, students deserve to hear why. And if they are simply ignorant of the extent of critique, well… that’s pretty damning of their professional credentials as supposed experts in the field of education, isn’t it? How am I to trust the scholarly judgement of my professors, their basic competence as researchers, if one of their key concepts is so mired in dispute? What are the alternative models?
Worse though, for all their talk, they never show the evidence. I am, at this point, in what everyone else calls “graduate school.” I completed a research methods-heavy social science program and graduated with two degrees. I and, I assume, my classmates can read academic literature. But we never do. Instead, professors waft about on the authority of “studies show” and “best practices” without ever presenting us that research. To move to an example beyond learning styles, my instructional methods professor says that inquiry-learning techniques improve student outcomes and that we should make every effort to incorporate inquiry into our classes.
Ok sure, but back up. Show me. Cite a study, demonstrate it, anything. Where is the evidence?
Given the uncritical embrace of learning styles, I suspect that it does not exist in a scientifically rigorous form (a pathetic 0.13% of studies in the field of education are replications …and we say psychology has a replication crisis!). Instead, most of our “readings” are blog posts, TED Talks, or, for lack of a better term, descriptive fact sheets. One of the readings began with “As an auditory learner…” I stopped right there.
Back in the day, my cosmology professor introduced us to another “dead theory,” that of “luminiferous aether.” Between Newton’s description of gravity in the 17th century and Einstein’s general relativity in the 20th, physicists explored the of concept of a static medium through which light might propagate, an “aether” or air that filled the universe. Relativity debunked the idea and, like epicycles, aether has become a term of derision. But my professor remained sympathetic to the idea, again, not necessarily as truth, but as an attractive metaphor for the constant dynamics playing out in “empty” space. Even Einstein himself, the man who killed aether, invoked the idea of a “new aether” in the 1920s to insist that space did indeed have “physical qualities.” No aether, so termed in its historic sense, exists. But neither is space a perfect vacuum. It’s full of “stuff,” a metaphorical aether. Just… smudge the details. It’s a pretty word for a pretty universe.
The rub here is that in light of the triumph of general relativity any modern invocation of aether must announce itself as a mere lay model or metaphor before digging into the grits of the quantum vacuum or whatever. “All models are wrong, some are useful…” – my professor used the idea to dispel the myth in us students that space was empty and to give the course a sense of historical arc from Aristotle and Plato’s elemental aether to modern quantum mechanics. There is some awe in the idea.
Compare that with the learning styles which my education professors embrace without apparent reservation. The theory could have value as a simple metaphor, as an unserious ice breaker or an opening to think about how students might differ. But instead, it becomes key to incorporate it into the differentiation of our instructional designs and obnoxious in its uncritical reoccurrence throughout my courses and readings. It has scientific pretentions, but we never seem to get to the science, which is weak regardless. “Evidence-based” education?
It isn’t even aether. It’s hot air.
(I’ve omitted one key point, which I intent to discuss in the next post: for all their talk of evidence-based practice, my professors seem incapable of assessing the science that produces the evidence for such a label)