In praise of signposts

//In praise of signposts

The safest road to hell is the gradual one – the gentle slope, soft underfoot, without sudden turnings, without milestones, without signposts.

C. S. Lewis

If you’re not sure which way to go, a sign post is very useful. A quick glance confirms either you’re headed in the right direction or you’re not.If you are facing in the right direction, all you have to do is keep on walking. Obviously you wouldn’t expect a signpost to contain much information about your destination; that’s not what they’re for.

This post makes the point that teachers rely on signposts to decide whether they’re teaching is informed by research but issues a warning:

Teachers don’t have much time to plough through complex research findings. What they want are summaries, signposts to point them in the right direction. But research is a work in progress. Findings are often not clear-cut but contradictory, inconclusive or ambiguous.

The post then examines some of the complexity buried in the research on learning styles. It makes the point that although most teachers have only heard about VAK, there are lots of other models out there and that on some measures some of them seem to perform better than others.

So maybe we should be a bit cautious about tarring all learning styles research with the same brush? Maybe there are some nuggets of purest gold buried amid the muck and filth? Maybe the signposts are a bit misleading?

Well, there are two points to consider here. First is the fact that as the author states, teachers’ exposure to learning styles is, in the main, limited to VAK. Certainly, I’d never come across any other style of learning style until a few years ago and I’ve never encountered anything other than VAK in schools. VAK, on the other hand, is pretty widespread. So any signpost which points away from VAK as being worthy of teachers’ time is helpful. What would be unhelpful is teachers being told, Hey! Some of the learning styles research which has nothing to do with anything you’ve ever been asked to do in school might actually suggest that some types of learning style aren’t all that bad. This would, I think, result in confusion and, inevitably, some people trudging off in the wrong direction.

The second problem is that while all learning styles might not be equally pointless, there’s enough evidence to suggest that they are not a profitable way for teachers to invest their time. Learning Styles: Concepts and Evidence by Harold Pashler, Mark McDaniel, Doug Rohrer, and Robert Bjork is my go to paper on this. It begins with a clear definition:

The term ‘‘learning styles’’ refers to the concept that individuals differ in regard to what mode of instruction or study is most effective for them. Proponents of learning-style assessment contend that optimal instruction requires diagnosing individuals’ learning style and tailoring instruction accordingly.

If you’re advocating something other than what’s encapsulated by this definition then you’re a lot better off calling it something else.

They then set out the criteria for determining whether any learning styles approach could be considered effective:

First, students must be divided into groups on the basis of their learning styles, and then students from each group must be randomly assigned to receive one of multiple instructional methods. Next, students must then sit for a final test that is the same for all students. Finally, in order to demonstrate that optimal learning requires that students receive instruction tailored to their putative learning style, the experiment must reveal a specific type of interaction between learning style and instructional method: Students with one learning style achieve the best educational outcome when given an instructional method that differs from the instructional method producing the best outcome for students with a different learning style. In other words, the instructional method that proves most effective for students with one learning style is not the most effective method for students with a different learning style.

This is clear, testable and, most importantly of all, falsifiable.

After combing through the research base, Pashler et al conclude:

… 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. [my emphasis]

So, does that mean learning styles are dead in the water? Well, no. Because the authors are scientists they admit “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.” But, the main finding is unambiguous:

The contrast between the enormous popularity of the learning-styles approach within education and the lack of credible evidence for its utility is, in our opinion, striking and disturbing. If classification of students’ learning styles has practical utility, it remains to be demonstrated.

Of course it’s true that “research is a work in progress” and beyond doubt that many findings are “contradictory, inconclusive or ambiguous”. As Pashler acknowledges, “Further research on the use of learning styles assessment in instruction may in some cases be warranted, but such research needs to be performed appropriately.”

As things stand if there was evidence supporting the use of learning styles I’m confident Pashler would have acknowledged and discussed it. The fact that they didn’t is all the support we need to continue warning teachers away from a seductive but ineffective approach to teaching and learning.


If you do think you’ve developed, or know of, an approach to learning styles which meets the conditions laid down by Pashler, you should get in touch with Will Thalheimer who is offering $5,000 to anyone who creates “a real-world learning intervention that takes learning styles into account – and proves that such an intervention produces better learning results than a non-learning-styles intervention.” Go for it!

2015-10-31T15:50:10+00:00October 31st, 2015|research|


  1. logicalincrementalism October 31, 2015 at 1:29 pm - Reply

    Some important points raised in this post David, but you are still viewing the evidence from the perspective of practitioners who generally want to know what they should use, not from the perspective of researchers who generally want to know what’s going on.

    Most school teachers might never have been asked to use a learning styles approach other than VAK, but I’m prepared to bet some of them have encountered Myers-Briggs during CPD, and Myers-Briggs according to Coffield et al scored reasonably well on reliability.

    I’m not suggesting signposts aren’t useful rules-of-thumb, but you’ve reinforced my suspicion that critics of learning styles are concerned mainly about their colleagues getting confused or misled than they are about the quality of research.

    In that case, why do learning styles critics place so much emphasis on the research?

    • David Didau October 31, 2015 at 1:38 pm - Reply

      You’re right – I am only really interested in the perspective of practitioners. Yes, my main concern is teachers being asked to waste time on unproductive approaches. We place emphasis on the research because this is one of the few cases where it is clear and unambiguous. The stage is set for researchers to prove the worth of their approach, why has no one managed it?

      As to Myers-Briggs, that’s had a pretty thorough mauling in recent years. Have you read Annie Murphy Paul’s book Personality? It provides a very down to earth overview of its flaws and limitations.

      Have you compared the Pashler review to Coffield’s ?

      • logicalincrementalism October 31, 2015 at 1:51 pm - Reply

        But the evidence isn’t clear and unambiguous. That’s the point of my post. Look at p.139 of the Coffield review.

        Researchers don’t ‘prove the worth of their approach’. They research. They often find evidence for and against a particular hypothesis and if they do, they are obliged to say so.

        I’m well aware of the problems surrounding the concept of personality. That makes the finding of reliability in respect of Myers-Briggs really interesting. *Something* is going on, but ‘personality’ might not explain it.

        I’ve read both Coffield and Pashler and noted their difference in approach. Will comment in more detail in a later post.

        • David Didau October 31, 2015 at 3:44 pm - Reply

          Is that the stuff on structures of power?

          I think we’re talking past each other: I’m more than happy for researcher to research learning styles if that’s what they want to do, but until they meet Pashler’s criteria I think I’m justified in advising teachers to steer clear.

          • logicalincrementalism October 31, 2015 at 4:02 pm

            No, it’s about the way personality (whatever that is) relates to how people perceive the world.

            You’re entitled to advise teachers to steer clear of using learning styles. What I don’t think you’re entitled to do is to describe the evidence as ‘clear and unambiguous’ when it’s not, on the basis of your opinion of the conclusion of a single go-to paper.

          • David Didau October 31, 2015 at 4:45 pm

            Right. I know you know that Pashler’s review considers many studies so there’s little point making that point, but unless you have a criticism of their test for learning style’s efficacy and the ‘meshing effect’ I’ll happily continue advising teachers to ignore learning styles.

            Also, although Pashler et al is my favourite source on LS, I have read others:

            Willingham, D. T., Hughes, E. M., & Dobolyi, D. G. (2015). The scientific status of learning styles theories. Teaching of Psychology, 42(3), 266-271.\

            Rohrer, D., & Pashler, H. (2012). Learning styles: Where’s the evidence? Medical Education, 46(7), 634-635.

            Klitmøller, J. (2015). Review of the methods and findings in the Dunn and Dunn learning styles model research on perceptual preferences. Nordic Psychology, 67(1), 2-26.

            There’s also this which albeit more cautious still supports my view that teachers are best off ignoring LS

            Furnham, A. (2012). Learning styles and approaches to learning. In K. R. Harris, S. Graham, T. Urdan, S. Graham, J. M. Royer, & M. Zeidner (Eds.), APA handbooks in psychology. APA educational psychology handbook, Vol. 2. Individual differences and cultural and contextual factors (pp. 59-81)

          • logicalincrementalism October 31, 2015 at 6:52 pm

            I’m well aware that there are a number of studies that conclude the evidence is insufficient to warrant teachers investing time and effort in learning styles approaches.

            But I’ll say it again. That conclusion doesn’t warrant making unsupported assumptions about constructs, theories, research methodology or what results show. If researchers are cautious in their conclusions, teachers should be too.

          • David Didau October 31, 2015 at 7:29 pm

            The conclusion that “the evidence is insufficient to warrant teachers investing time and effort in learning styles approaches” doesn’t warrant me advising teachers not to invest time & effort in learning styles?

          • logicalincrementalism October 31, 2015 at 7:32 pm

            That’s not what I said.

            You are perfectly justified in advising teachers in that way. What isn’t warranted is making unsubstantiated claims about the evidence, such as saying that it’s clear and unambiguous. If there’s one thing the review papers are in agreement on it’s that the evidence is patchy and confusing.

          • David Didau October 31, 2015 at 7:38 pm

            It is clear and unambiguous that there is no evidence to support teachers investing time or effort in learning styles. That is clearly substantiated. (Even by you in an earlier comment.)

            As I said in the blog, further research may be warranted.

            I really can’t see how you can disagree with that.

          • logicalincrementalism October 31, 2015 at 7:57 pm

            I think we’re using the term ‘evidence’ in different ways. I’m using it to mean ‘research findings’, you’re using it to mean ‘justification for practice’?

          • David Didau October 31, 2015 at 8:02 pm

            Yes, maybe that is the problem.

        • Michael November 4, 2015 at 6:05 pm - Reply

          Your being unreasonable with your criteria. Teachers are interested in the impact on the classroom. Both yourself and David agree that the evidence doesn’t support the use of VAK/learning styles in the classroom as they have not met the burden of proof. While understanding the implications of Myers-Briggs may be interesting its not directly relevant to educational practice at the moment You seem to be looking at this from a philosophical viewpoint rather then using the scientific method to unpick and resolve a specific question. (Will using learning styles in the classroom improve learning). Even a simple question like this can be difficult to keep our focus on. If we don’t stay focused none of our results will be useful as there will be to many possible variations to account for.

          Remember evidence based anything is supposed to place the onus on an intervention to prove its worth. A lack of clarity on an issue (for example regularly conflicting results which are unclear) is considered a lack of evidence to support that intervention. False positives and false negatives tell us that we should expect there to be patterns in data that support certain theories. This is why we are supposed to demand a strong consensus of repeatable data. .

          I release that more esoteric thinking can be useful but we live in a world were we want to use our understanding to improve our practices. Everything I have said above is implicit in Davids reasoning from the beginning of his post. You seem to recognize that but wish him to extend his definition to include wider ideas (Sorry if I am misunderstanding you here) but to me this seems dangerous close to the special pleading fallacy.

          • logicalincrementalism November 5, 2015 at 2:52 am

            Michael I think you’re using the word evidence in the same way as David Didau does – to refer to whether or not the evidence is sufficiently robust to support a particular intervention. As you say “A lack of clarity on an issue (for example regularly conflicting results which are unclear) is considered a lack of evidence to support that intervention”. I’m using the word evidence in a broader sense to refer to the results of research into learning styles in general.

            I understand that teachers are primarily interested in the impact of an intervention in the classroom, so their focus is likely to be on the conclusions that can be drawn from the research evidence rather than the research evidence per se. What has concerned me is that the conclusion drawn from the evidence (not worth using learning styles in the classroom) is being applied to the research evidence itself (if it’s not worth using learning styles then there can’t be any evidence to support their use, or, learning styles don’t exist).

            Unwary teachers might conclude that the research evidence clearly shows no support whatsoever for any kind of learning style intervention. That is not the case.

            If teachers are being encouraged to engage with research (reading it, doing it) then we need to be clear about what we mean by evidence.

          • Michael Pye November 5, 2015 at 6:00 pm

            Sorry I cant seem to but this after your post. I am using evidence the same way as David. It is true that evidence could be defined differently but I believe the definition we are using is the most widely accepted one in this field and more importantly the most relevant and appropriate to the situation. This could be wrong but I believe it is unlikely and i am using it as the arbiter of my reasoning. (Or as one of my premises). If you will indulge me an analogy your turning up to a rugby league match and criticizing them for not following union rules. (Or perhaps more accurately for not following Ruger rules as the approach you seem to be wanting to use is so much more philosophical then scientific).

            Incidentally unwary teachers (or anyone else) can misunderstand the evidence (or anything else) and in fact we are all in a state of mixed understanding/confusion. This proves nothing other then being a statement on the human condition. (Or thinking condition if you believe in aliens).

            Finally I don’t believe that evidence supports learning styles (any of them) as i have seen no evidence to refute this. (Note that this relies both on my exposure to the evidence, my understanding and finally my interruption of its validity). This is reasonable and logical irrespective of whether I change my mind down the line.

            Finally even if I find certain learning style theories interesting and fascinating, even if i find some evidence for their mechanics it does not mean there useful or correct.

            In physics there are some weird theories (see Alcubierre drive for one of my favorites) these theories often have plausible mechanisms and evidence to support their use however they have not been subjected to rigorous tests to determine their validly (usually because we lack a practical scientific design) as a result they are not currently falsifiable. As a result my dream of a warp drive is possible but not very likely (most theories are wrong so a reasonable person assume there ideas are equally unlikely to be correct ). You might like an idea, it may be extremely elegant and useful it is still likely to be wrong (or serious misunderstood) until rigorously proven otherwise.

            As an evidence based practitioner of teaching I am currently subsisting on a spartan diet. My reading (which is quite a lot but not nearly enough has told me the following)

            Learning styles don’t improve teaching (But do make teaching harder)
            Observations have poor objective accuracy (You need a good measuring device to measure something)
            A 1:1 teaching assistant for your child will hinder their learning (Yep that ones a shocker)
            Teaching assistants may help outcomes across a school. (And no if you read the research that’s not actually contradictory)
            I don’t know how to use teaching assistants more effectively
            (but i have ideas)
            I cant define learning accurately. (But I can assess if I have fulfilled my objectives).
            There is no evidence for the superiority of SMART targets over plain old targets.
            Micro-teaching may help students stay focused but there is no
            NLP and Brain Gym are pseudoscience.
            If i want a new idea for teaching looking at Coffields metanalyssis is a reasonable place to start. (Make sure you look at the latest ones they do change)
            No one can define dyslexia accurately and in a way that is repeatable and useful for the purpose of choosing an intervention (likely insert several other conditions here) and it may not even exist as a distinct entity of reading difficulties. (As a statemented Dyslexic this one was interested)

            Note that after 10 years in the profession and an inordinate amount of reading compared to my peers the above list is the best
            I can manage. If I use your definition I would have exactly zero in the above list. Every point above has fringe research debunking my conclusion (and yes I realise I could be wrong in my informed opinion).

            I know this was a long reply but I am relay hoping you will reconsider your definition of evidence.

            I would suggest the following

            Is there a reasonable mechanism explained
            Does the evidence predict occurrences accurately before they are recorded.
            Has the evidence been rigorously challenged looking for flaws.
            (This requires falsability which is why David is banging on about it)
            Have we defined what new evidence would be required for us to review our conclusion. (So we can recognise the need to challenge our biases)

            I think only the first two of the above have been met with learning styles. (yes all of them).

            Wonder if anyone bothered to read this far.

  2. joiningthedebate October 31, 2015 at 1:42 pm - Reply

    I’ve just had a brainwave (whatever that means). What usually happens is that senior managers believe in something for whatever reason (maybe they’ve paid a consultant). This then gets imposed on other staff who have to be seen to be following a whole school policy. Instead what if this happened… A senior manager believes in some new pedagogy. They then take responsibility for TEACHING a core subject for a whole year to a large class for example yr 9 set 2. At the end of the year when set 2 results are better than expected (or even better than set 1) we will all, how did you do it, let us copy your teaching methods. For primary schools I am sure something could be arranged for example the head taking on a parallel class for a term at least.
    The benefits of this approach are obvious:
    complaints from ordinary staff about senior hypocrasy are reduced;
    workload is reduced because it is only the senior manager who is trialling the new philosophy;
    if successful the evidence would be far more compelling because there would have been lessons that could be observed and students who could be interviewed, as well as resuls that could be analysed.
    Finally my suggestion fits in well with union guidelines (in my own words …) that new systems should be assessed in advance for the impact on workload and that new systems should not be introduced mid year.
    I don’t know why I don’t think of this sooner.

  3. joiningthedebate October 31, 2015 at 1:44 pm - Reply

    I don’t know why I didn’t think of this sooner (blame my tablet)

  4. edsacredprofane October 31, 2015 at 5:54 pm - Reply

    David aren’t you guilty of using the kind of closed arguments that you talked about in your last blog.

    You adopt a definition that agrees with you and then insist its the only definition you are prepared to accept.

    The field of learning styles includes cognitive styles, intellectual styles and, just, styles. It now concerns itself with meta-cognitive properties of different learning modalities.

    Of course I don’t think anyone will ever prove that delivering learning to a learning style will improve outcomes there are far too many variables involved. I wouldn’t ever recommend that teachers use learning styles for that reason.

    My guess is that feedback would have similar problems as would many other teaching and learning interventions that do not relate to a relationship between inputs and long term memory.

    It’s a kind of discursive game to set a high bar on evidence because inexorably you won’t be able to recommend anything other than what can be proved. Relying only on what can be proved may in itself not be a good basis for a teaching and learning approach

    I think the point that Sue is making is that dismissing research on the basis of weak evidence is as bad as accepting it on the same premise.

    You are in danger of creating another meme, which is that we ll learn the same.

    For example, here is Tm Oates appearing to endorse that meme and the view that we all learn in the same way. Albeit the quote is not exactly clear

  5. edsacredprofane October 31, 2015 at 6:12 pm - Reply

    I meant to say Tim Oates appears to be endorsing the fact that we all learn the same with the Dweck Meme.

  6. vlorbik October 31, 2015 at 6:45 pm - Reply

    /*A senior manager believes in some new pedagogy. They then take responsibility for TEACHING */

    i hope you’re joking. “TEACHING” is what
    they’ve been working for years to get *out*
    of. and that’s if we’re *lucky*.

    • joiningthedebate October 31, 2015 at 6:55 pm - Reply

      Oh yeah, sorry. I now see the flaw in my suggestion. Unless of course any senior manager out there wants to confirm my idea is workable…?

  7. teachwell October 31, 2015 at 10:19 pm - Reply

    Interesting comments section to say the least. Human history has been littered with bad ideas that have been abandoned if shown to be false. The learning styles advocates are an ‘end of the world’ cult who when it is clear that the world has not ended merely select another date (or in their case a different name for the same idea) and start again.

    • Brian November 1, 2015 at 10:19 am - Reply

      Teachwell … I used to work for a manager who actually used to say, “I have made my mind up, please don’t confuse me with the fact”. He didn’t even try to convince that he was right, just that his mind was made up and he wasn’t going to change it.

      You will also be aware that there are a few examples of ideas (bad simply confuses the issue) that are not abandoned even if not proved to be correct. In fact although I am neither a scientist or researcher my understanding is that much progess comes from this direction.

      To quote Plato….”one who listens to lumpers relentlessly on the twittersphere and bloggershpere tend towards lumperishness themselves”.

      It must be true as Plato said it.

      • teachwell November 1, 2015 at 11:13 am - Reply

        Brilliant – indeed I think some people’s capacity to cling to ideas is priceless. is not only a great example of this but also shows why ‘just google it’ will lead to many a problem in the future…

  8. Brian November 1, 2015 at 10:12 am - Reply

    This is what Pashler says……and indeed you quote in your blogpost

    “We conclude therefore, that at present, there is no adequate evidence base to justify incorporating learningstyles assessments into general educational practice.”

    Following the blogpost, this is what you say in your comment…….

    “the evidence is insufficient to warrant teachers investing time and effort in learning styles approaches”

    The second may well be taken from a different source, this is unclear, but clearly these two statements say different things. I am concerned that the first has morphed somehow into the second.

    I agree also with edsacred in that you choose “a definition” and then assert that this is “the definition”.

    I believe the issue of “lumping” is an interesting one. I believe that we all lump to a certain extent but it is when lumping is extrapolated beyond the data that I start to worry a bit.

    The Willingham stuff that I have seen is equivocal if it is read carefully and his infamous internet video refers to designing instruction to fit learning style and it’s “effectiveness”. Very rarely is it described thus in my experience.

    I have looked at my own learning style preferences and then fashioned my own autodidaxy (Candy) based upon the result. I am very happy with my learning. Not a scientific study by any means and I am not suggesting my results are generalisable but I do get the feeling that understanding an individual’s learning preferences may be useful and worth spending some resources on. For many kids autodidaxy may be (and probably is) their main route to learning.

    • teachwell November 1, 2015 at 11:17 am - Reply

      Ok not so sure where you are going with this. At some point the people who favour learning styles need to match Pashler et al and give evidence that it does improve outcomes instead of just saying it does. The problem is that these ideas are unfalsifiable as in the end if they don’t work or improve outcomes then the standard cry is ‘the teacher did not use learning styles properly’. If it is opinion against opinion then that is one thing but we are now in the realm of opinion vs research.

  9. […] October In praise of signposts – Rather than being worried about the fact that teachers don’t trawl though lengthy […]

  10. […] Didau responded to my thoughts about signposts and learning styles on his blog. Our discussion in the comments section revealed that he and I use the term ‘evidence’ […]

  11. Hugo Kerr November 2, 2015 at 7:41 am - Reply

    A thought: we know that students, especially weaker students, tend to follow learning styles taught to them – that the means of addressing learning are also learned. We need, then, to ask, when faced with a student who “has a learning style” whether it is learned, or whether it is innate. If it is innate, a real characteristic (I don’t think this possiblilty is proven, but if…), then we should, perhaps, teach to it. If, however, it is a learned style the student needs exposure to other styles rather urgently.

    A student who always “sounds out” may need visuo-motor approaches, for example, or vice versa. They definitely do if their style is a learned thing.

Constructive feedback is always appreciated

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: