Decades of research suggest that learning styles, or the belief that people learn better when they receive instruction in their dominant way of learning, may be one of the most pervasive myths about cognition. Nonetheless, little is known about what it means to believe in learning styles. The present investigation uses one theoretical framework–psychological essentialism–to explore the content and consistency of people’s learning style beliefs. Psychological essentialism is a belief that certain categories (such as dogs, girls, or visual learners) have an underlying reality or true nature that is biologically based and highly predictive of many other features. We tested the prevalence of erroneous essentialist beliefs regarding learning styles in both educators and noneducators, including that learning styles are innate, unchanging, discrete, and wired into the brain. In each of two experiments, we identified two groups of learning style believers, with one group holding an essentialist interpretation of learning styles, and the other group holding a nonessentialist interpretation of learning styles. No differences were found between educators’ and noneducators’ beliefs. In fact, only one factor was found to be a significant predictor of learning style beliefs for educators: the age of the population with whom they work. Specifically, those who worked with younger children were more likely to interpret learning styles in an essentialist way. Together the findings demonstrate that learning style beliefs are far more complex and variable than previously recognized.
Lots to chew on here. Two things in the above abstract jumped out at me:
1. “Psychological essentialism is a belief that certain categories (such as dogs, girls, or visual learners) have an underlying reality or true nature that is biologically based and highly predictive of many other features.” This belief is certainly true of dogs, girls, and lots of other categories! So I guess the point is that psychological essentialism, as defined above, is not in general a “myth.” The mistake is not in the belief of psychological essentialism, it’s in extending that belief too far (for example, thinking that dogs will always behave in stereotypically dog-like ways, or thinking that girls will not be interested in sports, or whatever).
If psychological essentialism were always wrong, I guess it wouldn’t be so much of a problem.
There are some always-wrong beliefs that many people have (for example, believing in ghosts), and there are some often-wrong beliefs that many people have (for example, believing in papers published in PNAS), but those erroneous beliefs can be walled off from the rest of our lives in many ways. For example, a belief in ghosts can add some spice to your life without otherwise affecting much of what you think and do, and you can enjoy podcasts that uncritically report on junk science without it getting in the way with your work.
Psychological essentialism is different. It’s a framework for sets of beliefs that are often correct and are also pretty much necessary for us to function as thinking beings. The very value of psychological essentialism helps explain how combating its overgeneralization is a major challenge; indeed, it’s something we’ve studied in our research.
2. “learning styles, or the belief that people learn better when they receive instruction in their dominant way of learning,” is described as a “myth.” I had no idea! It seems clear just from everyday life that different people learn in different ways, so I was stunned to learn that the professionals don’t think learning styles is a real thing.
But here’s what it says in the article:
Learning style philosophies are considered a myth because they provide anywhere from inadequate to incorrect portrayals of learning (Dembo & Howard, 2007; Pashler et al., 2008; Scott, 2010). To date, there has been no evidence that matching or meshing instruction to someone’s self-reported learning style positively affects their ability to learn new information (e.g., Husmann & O’Loughlin, 2018; Knoll, Otani, Skeel, & Van Horn, 2017; Krätzig & Arbuthnott, 2006; Rogowsky, Calhoun, & Tallal, 2015; also see Pashler et al., 2008 for a review).
Wow. Let’s check out Pashler et al., 2008, for a review. From the abstract:
Assessments of learning style typically ask people to evaluate what sort of information presentation they prefer (e.g., words versus pictures versus speech) and/or what kind of mental activity they find most engaging or congenial (e.g., analysis versus listening), although assessment instruments are extremely diverse. The most common—but not the only—hypothesis about the instructional relevance of learning styles is the meshing hypothesis, according to which instruction is best provided in a format that matches the preferences of the learner (e.g., for a “visual learner,” emphasizing visual presentation of information). The learning-styles view has acquired great influence within the education field, and is frequently encountered at levels ranging from kindergarten to graduate school. . . .
The authors of the present review were charged with determining whether these practices are supported by scientific evidence. . . .
Our review of the literature disclosed ample evidence that children and adults will, if asked, express preferences about how they prefer information to be presented to them. There is also plentiful evidence arguing that people differ in the degree to which they have some fairly specific aptitudes for different kinds of thinking and for processing different types of information. However, we found virtually no evidence for the interaction pattern mentioned above, which was judged to be a precondition for validating the educational applications of learning styles. Although the literature on learning styles is enormous, very few studies have even used an experimental methodology capable of testing the validity of learning styles applied to education. Moreover, of those that did use an appropriate method, several found results that flatly contradict the popular meshing hypothesis.
We conclude therefore, that at present, there is no adequate evidence base to justify incorporating learning-styles assessments into general educational practice. Thus, limited education resources would better be devoted to adopting other educational practices that have a strong evidence base, of which there are an increasing number. However, given the lack of methodologically sound studies of learning styles, it would be an error to conclude that all possible versions of learning styles have been tested and found wanting; many have simply not been tested at all.
Wow. Now I want to see the list of the “other educational practices that have a strong evidence base, of which there are an increasing number.” I’m not doubting that these exist, and maybe many of them are practices of which I’m already aware, such as active learning.
I’d just like to see that list, as it does seem that the education efforts I’ve observed, at levels ranging from kindergarten to graduate school, are not always evidence-based. I keep ending up on committees that are purportedly about improving education but end up involving long discussions of what should the rules be for bathroom breaks on exams.
Great find. Thank you. In terms of 1. I may be misunderstanding, but I think you might be conflating psychological essentialism with cognitive categorisation.
If so, it might help to say that a cognitive category is what transforms a set of stimuli into a recognised ‘thing’ with boundaries (however fuzzy) and features. Psychological essentialism is when a cognitive category is imbued with a theory (in the Douglas Medin sense) that the category is underpinned by some kind of defining core reality (a bit like the concept of a soul for an individual)
Cheers
Disclaimer: I have no specific training or expertise in psychology or human development. But I have extensive teaching experience with students and instructors who believe in learning styles. This commonly is expressed as a left brain/right brain dichotomy or a quantitative/qualitative mode of thinking. I’ve never thought these beliefs were evidence based (though my caveat above applies to my understanding). But I have found these beliefs to be counterproductive, even destructive. I frequently encounter students who fear and avoid quantitative analysis, citing their assertions regarding their biologically-based learning style. I’ve always doubted that and only saw it prevent their development of abilities I believe they possessed. I’d add that these beliefs seemed to be more strongly held by instructors than by the students themselves. I look forward to being better educated through the comments.
Quote from above: “I frequently encounter students who fear and avoid quantitative analysis, citing their assertions regarding their biologically-based learning style. I’ve always doubted that and only saw it prevent their development of abilities I believe they possessed.”
I think I suck at statistics, quantitative analyses, formulas, numbers, etc. I don’t think that’s strange, it’s only annoying because I needed to learn about that for my research master behavioural science, and before that even my bachelors degree in clinical psychology.
I wrote “this is easy” or something like that on my notebook for statistics classes, and passed all my exams during my research master, but still thought I didn’t really understand things. I heard about a book titled “Discovering statistics using SPSS” or something like that, which I read during the summer break on my own account just before my 2nd year where I needed to perform an experiment and analyze the data.
I still think I suck at statistics, quantitative analyses, formulas, numbers, etc. I severely doubted whether I should pursue a PhD because of that. I think it’s a good thing that I ended up not being a researcher in that sense, because I think I just don’t have the brain for that stuff. I’ve tried to make the most of what I think my brain is good at, so perhaps it all worked out pretty optimally.
You didn’t read the right book. Try:
R. J. Senter. 1969. Analysis of Data. Glenview IL: Scott, Foresman․
It’s an oldie but a goodie. Work your way through this book – and I promise that you can – and you’ll come away with a clear understanding of basic inference and ANOVA. With that behind you everything concerning regression and its uses (and disuses) is a lot clearer. I was pretty helpless around stats until I read this; now I’m a decent data analyst.
Yes, exactly, there is a problem where students and especially faculty/teachers believe that some people just aren’t able to do math. Of course dyscalculus exists for a small number of students. But I’ve actually read studies where teachers or parents saying things along the lines of “don’t worry, some people are just not good at math” or “I’m not good at math either” is harmful.
Dale –
I took a Montessori training course (in Bergamo, Italy – not exactly a hardship), way back in the day.
In that course, we were introduced to concrete materials to use in guiding students towards a more abstract mastery. The idea is that first you teach the underlying concepts and then you teach arithmetic algorithms (way for how to multiply two, two-digit numbers, or for how to cube trinomials). It was interesting that during the course, not a few of my fellow students said something on the order of:
“Oh, I always thought ‘I can’t do math,’ but now I see that the problem was that I was asked to memorize an algorithm that I didn’t really understand. Looking around I saw that other kids were able to master the algorithm, and so I felt stupid, and so then I decided that ‘I can’t do math.’ Going back to understand the underlying concepts, concretely, I see that actually, I can to math.”
Here’s my take, broadly. There’s a natural developmental (learning) sequence that applies generally. That sequence is that you move from a solid foundation concrete understanding to a more abstract understanding. In schools, at least in previous decades, there was an insufficient focus on that sequences. Either, (1) there wasn’t enough focus generally on exploring the “concrete” part or (2), individual students were pushed into a more abstract domain before they, individually had sufficiently explored the concrete (conceptual) domain. This is connected to the underlying paradigm whereby students are judged and evaluated on the basis of how quickly they move through the scope and sequence of a non-individualized curriculum.
I connect this with the idea of “learning styles.” Perhaps being a “visual learner” translates as “a learner who really wants to understand something deeply and feel comfortable with it before working in a more ‘abstract’ domain. This suggests to me that the concept of “learning styles” is on to something, but like many aspects of education, it becomes a matter of “validity” (as in testing) where you’re measuring a phenomenon by counting with a metric that doesn’t actually measure what you’re trying to measure.
You offer a possible interpretation, but I will propose a different one. There are surely different ways of teaching and some are better than others. “Teaching style” might be what you are observing, not “learning style.” Quantitative subjects are often taught badly – the fact that some people can still learn from bad pedagogy says something about their innate talent, but I’m not sure it really represents a learning style. I read your story as indicating that good pedagogy can override supposed learning styles.
Dale –
That’s a good point. I agree pedagogy is a key factor. But I think it’s not a dichotomy; pedagogy (teacher-centric) and “learning style” (student-centric) interact in complex ways and effective teaching adapts accordingly.
For instance, students benefit from differing amounts of work with concrete materials before tackling abstract concepts. But I think that it’s more that just talent that describes this. But could stem from life experiences or traits like confidence or curiosity, which don’t always align directly with talent. I have a theory that kids who tend to want to understand the reason behind a task may benefit from more concrete exploration, while those who want to be told what to do might grasp abstract algorithms more easily.
Still, I agree talent plays a role, and pedagogical style interacts with individual student characteristics (lumped together under the vague descriptor of “learning style,” which blends variables like talent, curiosity, confidence, experience, and preferences). This connects with the references in some of the other comments to the importance of appropriate pacing.
As someone who:
-was generally overwhelmed at high school,
-couldn’t concentrate while having someone else in class read while I was also supposed to read the same thing at the same time,
-got low grades in high school chemistry except for that one time I asked the teacher some stuff after class,
-dropped out of high school, did some odd jobs while studying at home by myself and got a diploma that way,
-did some odd jobs again, and traveled a bit, before entering university at a relatively later age where I skipped a lot of large scale classes during my bachelors degree
-entered a research master and found it very useful to be able to ask questions in class immediately because of the small classes
-even at the research master noticed that I couldn’t write or type (except very brief) while listening to the teacher at the same time
I definitely think there are differences in how people learn.
I think we can all agree that individuals differ in how they learn (I know I believe that). But I think the question here is about learning styles across “categories” of people. I think these are group-based differences based on biological characteristics (e.g. sex, race, age, ethnicity). Is the variability across these groups large relative to the variability across individuals within these groups? Are these group differences predictable? I don’t believe that is the case, although I admit to not having expertise about this.
Quote from above: “But I think the question here is about learning styles across “categories” of people.”
I can imagine my personal characteristics and experiences might be tied to some sort of category (e.g. being on the autism spectrum might be an option here). And I can imagine that several of the things I mentioned might be the case for other people (some of which might be on the autism spectrum, whatever that term is worth or means or implies).
I don’t really like working in group settings, or doing two things at the same time (e.g. writing and listening). I can more likely focus very well on a single thing at a time. I don’t do well transitioning between tasks or subjects, etc. etc. I think these sort of things might all be things many more people might also experience or find.
For me personally, I think high school and universities and academia are for a significantly large part just not made for my brain. It just doesn’t work well with how I learn and function optimally, and with what I think is important in education. It’s all just not how I think things should work, or how my brain works. This is probably why I dropped out of high school, and started university at a later age, and also not really functioned there.
I am at the moment pretty happy though with my attempt to make the most of my brain in light of learning and education and all the problematic issues for me regarding those things. I even did some writing and publishing, and engaged in online discussions, etc. I think I made the most of what I was given in that regard. At least I hope I did.
Dale, you say “I think we can all agree that individuals differ in how they learn (I know I believe that). But I think the question here is about learning styles across “categories” of people.” But the articles Andrew quotes are talking about individuals, not categories. For instance, “… there has been no evidence that matching or meshing instruction to someone’s self-reported learning style positively affects their ability to learn new information”, and “The most common…hypothesis about the instructional relevance of learning styles is the meshing hypothesis, according to which instruction is best provided in a format that matches the preferences of the learner.” There’s nothing there about categories; both of these explicitly refer to the preferences of individual learners.
That said, I do agree with you that individuals differ in how they learn. That’s the puzzle! If we differ in how we learn, how come these studies say we don’t?
To put it another way: this is one of the many, many situations in which the true effect can’t be zero — the optimal way of teaching someone can’t be exactly the same for everyone — so we are really talking about how big the difference is…and how this might be tested, and many other issues. For instance, the way people prefer to be taught might not turn out to be the optimal way.
I suspect the effect size is not huge, the methods for measuring it are noisy, the sample sizes of the individual experiments are small, and the experiments are not all well-conceived or well-conducted.
Phil:
You ask, “If we differ in how we learn, how come these studies say we don’t?” But I don’t think the studies say that people don’t differ in how they learn. Rather, the studies say that the approach doesn’t work, of teaching in this particular way in response to these apparent learning styles. The thing that doesn’t work is using teaching method X on people with learning style A, using teaching method Y on people with learning style B, etc. The problem is, first, that the learning style assessment is noisy; second, actual differences in learning style don’t correspond so clearly with A, B, etc.; third, the recommended teaching approaches are not necessarily working in the way they are attended.
To put it another way, to have a better teaching approach it’s not enough to know that people learn in different ways. A lot of details need to work too, and it’s easy for people to try out educational strategies that seem logical but don’t really do much. Hence this sort of study.
Those quotes about “self-reported learning style” I take to mean that they self identify with some category, such as ‘visual learner’ or ‘left-brain.” So, I believe they are referring to categories. If not, then I am wrong – but I’m not sure how research can be conducted about things like “matching or meshing instruction to someone’s self-reporting learning style.” I guess you could compare a treatment (matching or meshing) to a control group where no attempt is made to do that. But I think such research would be doomed from the start. At best, it might say something about a particular attempt to match or mesh done by a particular person in a particular environment. But I doubt there would be any general conclusions about the effectiveness of trying to match. On the other hand, if the self-identified styles are categories, I can more easily imagine how such things can be studied.
Dale, you said earlier “I think the question here is about learning styles across “categories” of people. I think these are group-based differences based on biological characteristics (e.g. sex, race, age, ethnicity).” But that is not what these studies studied; I think we agree that they separated people based on self-identified learning style…which, I agree with you, can also be called different ‘categories’, but perhaps you can see why I was confused about what you were saying.
Doesn’t matter…as for me, my main point stands: I suspect the effect size is not huge, the methods for measuring it are noisy, the sample sizes of the individual experiments are small, and the experiments are not all well-conceived or well-conducted.
Hi Anon,
There are certainly differences in perception, cognition, emotion and personality* between people, and these differences in how we perceive, think about, and behave in, the world indeed affect how (and what) we learn.
However, although I’m not an expert, I think ”learning styles” usually refers to something much more categorical and simplistic. For example, as a primary school teacher, I might divide my class into ”tactile”, ”visual” and ”auditive” learners. I’ll then have the tactile learners learn addition and subtraction by moving wooden cubes around, the visual learners will watch someone demonstrate how to do it on the board, and the auditive learners will be told how to do it.
There is no scientific basis for that kind of approach.
So while we are all different, and a good teacher should try to adapt their teaching depending on the students, the specific didactic approaches that are called ”learning styles” are probably not helpful.
Forgot to spell out my *:
*One could ask whether it makes sense to list personality as a ”thing” in itself (in the manner I just did) or whether personality is an emergent property of differences in perception, cognition, emotion, et cetera. But that is for another day.
That was the point I was making, but your statement is better than mine. Grouping people by “learning styles” might be biologically based (I am familiar with the old, erroneous, version that women and men differ in terms of right brain/left brain styles), but they can also be based on perceived groups. A teacher that separates students on the basis of a perceived learning style (e.g. visual vs written or spoken) might not have an biological basis, but it still presumes that there is a valid difference between such groups.
I wonder how this might relate to issues of marketing. Marketers typically classify consumers into groups (clusters). There is individual variability within these groups, but there is a belief that the between group differences are still relatively large, stable, and/or predictable. Are such preferences qualitatively different than differences in learning styles? If we believe the former have a sound empirical basis, does that imply anything about the latter?
As a former high school teacher, I can confirm that that is exactly how “learning styles” is understood.
Correct, and then teachers were told they had to come up with 4 different ways to teach each piece of content to mesh with the 4 different learning styles. It was a ridiculous piece of use of questionable research for policy implementation.
Even Dog is not all that essentialist. I mean, you can breed dogs with wolves, coyotes, golden “jackals”, but not with “African wild dogs” which are not Canis nor with true Jackals. But a King Charles Spaniel is obviously very different from a German Shepherd or a Cane Corso even though they’re all domestic dogs. There are of course many similarities as well.
The point of “essentialism” if I understand it, is that there are some features which simply can not vary. When you look carefully even some obvious ones are wrong. All dogs have four legs, except there are many tripod dogs or dogs with back wheel prosthetics. All dogs have hair, except hairless ones like Xoloitzcuintli. Dogs have tails, except the ones that are docked. Dogs bark, except the ones that dont. Dogs can’t fly, but there are military dogs that jump from airplanes and parachute…
With biology the rule is variability. There are no dogs that lay eggs, or have scales, or feathers, or photosynthesize though.
Andrew says:
Wow. Now I want to see the list of the “other educational practices that have a strong evidence base, of which there are an increasing number.”
If I’m not wrong, the following are some of them:
Retrieval Practice: Actively recalling information boosts long-term learning (Roediger & Karpicke, 2006; Karpicke & Blunt, 2011).
Spaced Practice: Spacing study sessions over time improves memory retention (Ebbinghaus, 1885; Cepeda et al., 2006).
Interleaving: Mixing different topics or problems during practice enhances learning and transfer (Rohrer & Taylor, 2007; Pan, 2021).
Thanks!
A colleague in my department works on this topic. Here’s a book chapter describing some general ideas:
https://learninglab.psych.purdue.edu/downloads/2024/2024_Karpicke_ODay_Oxford_Handbook.pdf
Here are some specific ideas that have been investigated and a few of his studies (not always showing a positive effect):
Judgments of learning
https://learninglab.psych.purdue.edu/downloads/2021/2021_Ariel_Karpicke_Witherby_Tauber_EDPR.pdf
Concept mapping
https://learninglab.psych.purdue.edu/downloads/2021/2021_ODay_Karpicke_JEDP.pdf
https://learninglab.psych.purdue.edu/downloads/2014/2014_Blunt_Karpicke_JEDP.pdf
Video-based learning:
https://us.sagepub.com/sites/default/files/hevideolearning.pdf
Retrieval practice:
https://www.frontiersin.org/journals/psychology/articles/10.3389/fpsyg.2016.00350/full
https://learninglab.psych.purdue.edu/downloads/2014/2014_Grimaldi_Karpicke_JEDP.pdf
The only mention of evidence backed approaches in the actual paper (only skimmed) was this:
“For example, the finding that learners’ memory for information or procedures can be directly enhanced through testing (Roediger & Karpicke, 2006) is not something that applies to only a small subset of learners but (as far as can be told) applies to all. Although performance of a student on a test will typically depend on that student’s existing knowledge, testing (when carried out appropriately, which sometimes requires providing feedback) appears to enhance learning at every level of prior knowledge.”
following up on this request: Now I want to see the list of the “other educational practices that have a strong evidence base, of which there are an increasing number.”
This is a good review of effective educational practices with solid empirical basis:
Dunlosky, J., Rawson, K. A., Marsh, E. J., Nathan, M. J., & Willingham, D. T. (2013). Improving Students’ Learning With Effective Learning Techniques: Promising Directions From Cognitive and Educational Psychology. Psychological Science in the Public Interest, 14(1), 4-58. https://doi-org.ezproxy2.library.colostate.edu/10.1177/1529100612453266 (Original work published 2013)
John Hattie published a synthesis of many metaanalyses on educational interventions. His results are often displayed in infographics illustrating the effect sizes:
https://visible-learning.org/2023/01/visible-learning-the-sequel-2023/
https://visible-learning.org/2016/04/hattie-ranking-backup-of-138-effects/
In general things that work typically aren’t cheap or easy (no such thing as a free lunch).
Hattie list:
https://www.visiblelearningmetax.com/influences
https://www.visiblelearningmetax.com/influences/view/matching_teaching_to_style_of_learning
I believe the What Works Clearing house is (or used to be) considered a good source.
https://ies.ed.gov/ncee/wwc/
In the “Learning Styles” framework you are supposedly able to determine you learning style through a questionnaire like this one: https://learningstyles.webtools.ncsu.edu/.
Lots of us attended professional development based on the idea that students can be classified into one of 3 or 4 groups and that they probably should only be taught using techniques designed for those groups. For every topic you had to come up with auditory, tactile and visual lessons. (No they cannot learn to use other styles; this is what is essentialist about the model.) (This is like Myers-Briggs, you *are* in a group; you cannot escape that group.)
Now, did this result in a lot of development of active learning lessons, absolutely it did. So that was an unintended positive consequence.
The summary at WIkipedia is pretty good. https://en.wikipedia.org/wiki/Learning_styles
Now, does that mean that everyone learns in the same way or has the same preferences? Absolutely not. For example, some people have very specific disabilities. But not only that. For example, I like to take notes. I like to draw pictures showing the relationships between things. That’s all fine and real, and it’s worth it for me to use what works for me. But that doesn’t mean a teacher has to spend their time adjusting to my preferences. Which are just that, not some biological characteristic.
This is all kind of odd. My memory has it that the learning styles idea had been convincingly rejected, in particular that (a) students themselves were really bad at identifying how they learned best and (b) that teaching in styles effective for the material being taught is a way better idea than teaching to the student’s hypothetical (and probably wrongly identified) “learning style”.
But I don’t think the “learning styles” idea is a “myth”: it’s a reasonable idea that happens to be wrong. Students are different, and a good teacher will have noticed that. And have looked to psychology for help/advice, and the early work supported the idea.
Also, while I love Wikipedia, it has the problem that it can’t say “this whole branch of work is friggin ridiculous”. So the article on B.F. Skinner makes his ideas seem like reasonable academic research. And the learning styles article is all “several studies have disputed this idea”. Sheesh.
> (b) that teaching in styles effective for the material being taught is a
> way better idea than teaching to the student’s hypothetical (and probably
> wrongly identified) “learning style”.
That’s my prior: The subject matter determines the way to teach the material, not the individual student. Since there doesn’t appear to be evidence to contradict my prior, I’m sticking with it.
People have their stated preferences. Also, they can learn certain things with relative ease or difficulty, say, language, music, math, drawing, the rote retention of information, abstract thinking. It seems to me that learning a particular skill has to be tied to the part of the brain responsible for that activity, so music has to be learned musically, etc… There will be variation in the preferences of individuals learning a particular thing, but does it really make sense to say people learn differently? Does it make sense to do an end-run around a skill and have people try to learn drawing musically, or math kinetically? Wouldn’t that end up delaying learning rather than accelerating it?
https://link.gale.com/apps/doc/A116184179/AONE?u=nysl_oweb&sid=googleScholar&xid=58ccb69e