Understanding Learning Styles Research

Too many people have been talking about learning styles research lately for me to try to cite them all here. Many have commented on the Learning Styles Don’t Exist video, for example. Via Karyn Romeis and Stephen Downes, I found two lengthy reviews of learning styles research:

These are two related reports with some shared content. The first is a literature review of 13 often-cited learning styles; the second summarizes that research and focuses more on how it can inform practice.

There’s a lot here, and it’s all still percolating in my brain, but I wanted to record some ideas from the second document.

Consistency, Reliability, and Validity

  • Much of the learning styles research is weak, failing on external reviews of consistency, reliability, and validity. Some models are better than others though.
  • Of the 13 models reviewed, only 1 passed all the psychometric measures: the Allinson and Hayes Cognitive Style Index.
  • Two models reviewed passed 3 of the 4 measures: Apter & Vermunt
  • Dunn and Dunn (the VATK model which is probably the most familiar in the US) passed an external review for predictive validity but not internal consistency, test-retest reliability, or construct validity.

Context

  • Research is sometimes generalized to contexts inappropriately. Even if something works in higher ed, it might not work in corporate training or vice versa.
  • Learning style inventories don’t always transfer well across cultures or socioeconomic classes; they may not fare well in translation.
  • There isn’t one universal model of “learning styles” that ties everything together, and the researchers don’t agree with each other.

Application of Research

  • The research that exists is often misapplied.
  • Claims of what research shows and the effects of using specific methods are “overblown.”
  • Even when learning styles are shown to have a significant effect, that effect size isn’t as large as some other strategies. Teaching metacognition skills, for example, has a greater effect. So does formative assessment.
  • Researchers and theorists don’t agree whether it’s better to match teaching style to learning style or not–the research has been inconclusive.
  • Using any learning style model makes people too eager to label each other.

What Next?

So what am I going to do about it? I admit that I accepted the idea of learning styles unquestioningly when I got my education degree. There’s certainly a common sense feel to some of the systems. The Dunn and Dunn VATK model has become sort of “folk wisdom” among teachers and even the general population. But is that really enough for me to use it in my own instructional design now?

I’ve got a couple of ideas of what to do next. These are my own learning goals, so to speak.

  • Keep reading and following the conversations. There’s a lot going on in multiple different directions, but I need to try to pull some threads out and make sense of them for myself.
  • Be skeptical of everything I’m reading in the conversations. There’s too much misinformation, misunderstanding, and misapplication. What sounds intuitively “right” may not actually be true, or at least not useful.
  • Learn more about the Allinson and Hayes model (Cognitive Style Index) mentioned in the literature review.
  • Don’t worry about this too much. In terms of instructional design, learning styles are not the most important thing for me to think about. I can focus my energies on a lot of other things without stressing too much about adhering to a specific model.

I’m curious what others are doing. Are these conversations on learning styles affecting what you’re designing right now? Or have you only strengthened your resolve for your prior position?

12 thoughts on “Understanding Learning Styles Research

  1. Of course, people don’t understand statistics either–and I’ll be the first to admit I could use some improvement in that area myself. I didn’t get Cognitive Daily’s error bar test right, after all. I get enough of it to recognize when the media describe presidential polls as saying something they don’t actually say though, which likely puts me ahead of a lot of the US population.

    Glad to see you’re having fun with Chrome. I like my Firefox add-ons too much to be looking for something else. 🙂

  2. Kia ora Christy!

    I agree that classrooms don’t behave like controlled environments. This is where statistics becomes useful and important. When gathering educational data, statistical considerations can permit us to decide what is significant and what is not. It can also define suitable parameters within which useful data gathering can be accomplished so that there is more likelihood of the data being significant once obtained.

    I’ve studied quantitative and qualitative research. I’ve also done some action research. Educationalists seem to use action research a lot. The statistical significance of action research is rarely high enough to be useful (often it is not at all statistically significant) and it is also rarely transferable with any usefulness.

    I’ve often thought it a pity that the term ‘action research’ was ever invented.

    My experience is that some educators misunderstand the transferability of action research findings and think that useful action research results, once completed, can be rolled out to a wider team. It is extremely rare that anything useful about action research can be rolled out in this way. Any attempt to do this is neither statistically significant, nor is it likely to be any use (even if it was).

    Ka kite

  3. There’s a big difference between “research” like a school research paper and academic research or controlled scientific experiments.

    The controlled scientific experiments have their place, even in education, but you’re right that they aren’t necessarily useful. Classrooms don’t actually behave like controlled environments, and what can work in a lab doesn’t always work with real students. There’s a vast gulf between the research being done and what teachers and instructional designers can actually use and do.

    In one of my college music history courses, we read Stephen Jay Gould’s “The Case of the Creeping Fox Terrier Clone.” It has nothing to do with music history, as you might guess, but everything to do with being skeptical of sources even when they are published and recognized. It was a great lesson in information literacy, and it really opened my eyes.

    What I’d love is to figure out how to give other people moments like I had with the Gould essay, something that would make people start to question the assumptions about supposedly reliable sources and what we really know. I haven’t got a clue how to do that, but I’m looking.

    At least I do feel I’ve done my small part recently in developing a course on 21st century literacy skills. One week of that course is on information literacy as a skill that both teachers and students need to be successful. Participants will be having conversations about info literacy and why it matters; even that is a place to start.

  4. Kia ora Christy!

    “a segue to the importance of information literacy and why we should teach people how to understand and evaluate research” – I couldn’t agree more!

    The word ‘research’ is being misinterpreted at all levels. Quite frankly, it is one of these words that has evolved to the point where it no longer means research – daft as that may seem. I know too many people who think research is simply reading up on a topic, drawing conclusions about it, and then implementing a strategy built out of the conclusions – and I blame, in part, the current methods of teaching that are used in primary and secondary schools today.

    Children are told they are doing research when they are given a project. That’s fine to a point. But when they leave school and go on through their lives thinking that this is what research is all about, I can fully understand why we have the problems that you and I are discussing here.

    Education research (funny enough) is in a poor state in many western countries. The problem as I see it, is in the method as well as in the understanding of what represents statistically significant information that might have been gathered from an attempt at doing education research. The current culture of regarding the usefulness of the so-called scientific approach also lends itself to this problem.

    “Information literacy” plays a big part here. It starts small with the understanding of the meanings of simple terms such as ‘about’, ‘approximately’ and ‘significant’.

    Cutting short what could be a very long comment here 🙂 we (society) need to take stock of how we got here, instead of always looking for a way forward. To coin a metaphor – if one gets lost in the forest, which way is forward?

    Ka kite
    from Middle-earth

  5. “What I feel strongly about is the all-or-nothing approach that is often adopted exclusively when some new pedagogical theory is launched with little research done.”

    That’s the crux of the whole matter, isn’t it? Learning styles are still often viewed as sacred and untouchable in the US. I think we’ve pushed what little research there is farther than it was meant to go.

    I’m not convinced that there’s nothing of value in the idea of learning styles. Clearly, individuals do learn differently, and we may be able to identify patterns in learning that constitute styles or strategies. Actually, “strategies” may be more the way to think of it–strategies that may be preferred by individuals, but can change depending on the context or the task or just because the learner is bored. At least in online learning, it seems easier for me to make all the options available to everyone and let people choose themselves, rather than denying anyone a resource because they have the “wrong learning style” for it.

    It’s easy to get caught up in learning styles or digital native/digital immigrant or brain-based learning or whatever the current fad is. But I think you hit the nail on the head about not taking an all-or-nothing approach to any of them.

    Seems like a nice segue to the importance of information literacy and why we should teach people how to understand and evaluate research better…

  6. Kia ora Christy!

    Considering that the jury has been out on learning styles for some time now, I am really surprised that anyone has majorly incorporated strategies that hinge on the so-called learning styles approach. For one thing, many major shifts in pedagogical approach that have been dabbled in over the last 20 years have not been too productive with respect to successful results. A typical example of this was when so-called ‘word recognition’ was applied willy-nilly to reading programmes in New Zealand. It has since been shown that this spelt disaster for some learners who would have benefited from phonetics but were denied that assistance because of the swing of the pedagogical pendulum.

    One of those learners was my daughter who, at 9 years old, had a reading difficulty. It took me, as an educated father (not a teacher of English nor reading, but a Science/Maths graduate) to identify that she could be assisted, out of sight, by using a phonetics approach with her reading. She works in 3rd level computer support now.

    What I feel strongly about is the all-or-nothing approach that is often adopted exclusively when some new pedagogical theory is launched with little research done. One of those ideas that has taken off, like a rocket with a short fuse, is the digital immigrant – digital native myth. The damage that this has done (and is continuing to do) to institutions that have adopted strategies based on this premise is appalling. The most recent researches are clearly indicating that there has been no dramatic change in the so-called ‘wiring of the brain’ for young learners that would make any significant difference at all in this regard – and the evidence can be found elsewhere too. One is in the statistical evidence coming to light about the use of mobile phones while driving and the accidents caused by drivers of all ages.

    Not getting to far off topic, my hunch is that learning styles, though it has been a fashionable ra ra idea, really has the potential to cause more learning problems in the long run than it has potential to assist through the possible explicit denial of means to learn to learners who have been ‘diagnosed’ as having a particular ‘learning style’.

    History has shown that we humans are slow to learn from past mistakes. We seem to be more interested in trying something newfangled method, and for that reason alone, than we are in trying something known to work. Tested methods are old hat. So much so that they get biffed without thought to why they were in place to start with.

    This is not just limited to teaching and learning either. The leaky building syndrome in New Zealand is a typical example of how this has happened during the past decade or so in the building industry.

    Oh dear! Oh dear! Oh deary dear dear!

    Ka kite
    from Middle-earth

Leave a Reply