Why This Article Exists
If you’ve read the other pieces in our series on recent research, you’ll have noticed something: alongside genuinely useful findings, we’ve sometimes had to add a caveat about how a study measured VARK preferences in the first place. That wasn’t an accident, and it isn’t rare. It reflects a real pattern in the research literature — one that anyone reading about VARK, and anyone planning to research it, should know about.
This isn’t a piece about whether VARK “works.” It’s about something more basic: whether a given study actually measured what it claims to measure.
A Surprisingly Common Problem
VARK is a popular topic for student research projects, dissertations, and journal articles — which is a good thing. But popularity doesn’t guarantee quality, and VARK has a specific, recurring vulnerability: it’s easy to look like you’re using it correctly while actually measuring something else entirely.
The most common issue is this: the VARK Questionnaire isn’t scored by simply tallying which modality gets the most answers and calling that the result. The official scoring algorithm accounts for how close the scores are to each other, and identifies multimodal profiles — bimodal, trimodal, quadmodal — rather than forcing everyone into a single category. A student who scores 7-6-6-5 across the four modalities is meaningfully different from one who scores 12-2-1-1, even though both might have “Visual” as their top score. Treating them the same — by taking the highest score and assigning a single label — produces a completely different, and much less accurate, set of results than the real questionnaire would.
When we went looking for research to feature in this series, this was by far the most frequent issue we ran into. Some studies simply took the top-scoring modality and stopped there. Others used adapted or shortened versions of the questionnaire without validating them. A few described the modalities in ways that didn’t match Fleming’s original definitions at all — for instance, describing “Kinesthetic” as purely physical movement rather than learning grounded in real experience and application, which changes what’s actually being tested.
None of this is necessarily done in bad faith. VARK looks deceptively simple from the outside — four categories, a short questionnaire — and it’s easy to underestimate how much the scoring nuance matters. But the effect on the research is the same either way: the published “VARK results” may not be VARK results at all.
What This Means If You’re Reading VARK Research
A few habits go a long way:
Check how preferences were determined. Look for a methods section that mentions the official VARK scoring algorithm, or at least describes multimodal categories (bimodal, trimodal, quadmodal) alongside unimodal ones. If a study only reports single categories like “40% Visual, 30% Aural…” with no mention of multimodality, that’s worth a second look — real VARK samples are usually mostly multimodal.
Watch for modality descriptions that feel off. If a paper describes Visual learning as “watching videos” or Kinesthetic as “moving around the room,” it’s likely working from a simplified or second-hand understanding of the model rather than VARK’s actual definitions. That doesn’t automatically invalidate every finding in the paper, but it’s a signal to read the rest more carefully.
Consider the source. Preprints, student dissertations, and articles in low-oversight journals can absolutely contain good research — but they haven’t been through the same scrutiny as an established peer-reviewed journal, and it’s worth checking who published a study and how it was reviewed before treating its numbers as settled fact.
Separate the idea from the data. Sometimes a study has an interesting, plausible idea — that active engagement matters more than modality, say, or that mismatched teaching styles can build adaptability — even when its VARK measurement is shaky. The idea can still be worth thinking about. Just hold the specific numbers more loosely than the underlying insight.
None of this means ignoring research that isn’t perfect. It means reading it the way you’d read any single study on a contested topic: as one data point, not a verdict.
A Note for Prospective Researchers
If you’re planning a study involving VARK, a little care early on saves a lot of trouble later:
Request permission before you begin. The VARK Questionnaire is copyrighted, and researchers are required to apply for permission to use it before starting their study. Reach out early — this also opens the door to getting feedback on your design before any data collection happens, which is the best time to catch problems.
Use the questionnaire exactly as provided, and score it with the official algorithm. Modifying or adapting the questionnaire itself — shortening it, rewording items, changing the response format — isn’t permitted. Neither is scoring it with anything other than the official VARK algorithm. If your research context seems to call for changes, that’s exactly the kind of thing to raise with us directly rather than deciding on your own; we may be able to point you toward an existing validated approach, or advise on how to proceed.
Report multimodality, not just single categories. If your results show everyone neatly sorted into four single categories with no bimodal, trimodal, or four-part preferences at all, that’s a sign something went wrong in scoring — genuine VARK samples are usually predominantly multimodal.
Get the definitions right. Before designing materials or writing up your discussion, revisit VARK’s descriptions of the four modalities rather than relying on second-hand summaries, which have a tendency to drift over time and across papers.
Talk to us. We’re happy to discuss your research plans and offer feedback or advice at any stage — not just at the permissions stage, but as questions come up during design, data collection, or analysis. Researchers who reach out tend to produce results the rest of us can actually rely on, which benefits everyone working in this space, including you.
Good VARK research is out there, and it’s genuinely valuable — studies that use the real algorithm, report multimodality honestly, and engage seriously with what the modalities mean can tell us a great deal about how awareness and reflection shape learning. But it takes more digging to find than the volume of VARK-related publishing would suggest, and that gap is worth naming rather than papering over.
