Using VARK in Research

VARK is a very active topic for researchers and this statement is an attempt to assist researchers from following unhelpful pathways.
Firstly, VARK is only part of a learning style. A complete learning style should have information on some 20-plus preferences that affect learning. This includes such things as preferences for learning at different times in the day. A learning style would be a complex amalgam of questions and analyses. VARK is an intensive look at only one set of preferences within a learning style – the preference(s) that learners have for their communication modes – Visual, Auditory, Read/write and Kinesthetic. VARK is copyright to VARK Learn Ltd, a registered New Zealand company  and VARK is also trademarked.


The following are some of the pitfalls that result in poor research and invalid results using VARK data.

1     In the seminal research by Walter Leite (see website reference) the arithmetic Standard algorithm used on the website was deemed inadequate for research purposes especially those that plan to be published.   As a consequence a VARK Research algorithm using standard deviations was prepared and is recommended for researchers.  We decline to grant copyright release for those who want to use paper copies of the VARK questionnaire because the participants receive no help from completing the paper version of the questionnaire and merely knowing their four VARK numbers.  VARK was primarily designed to help learners, not for research. 

2     Any hypothesis that attempts to find links, especially correlation significance, between VARK and academic success will be invalid and a waste of research time and money.  Academic success, is, of course, poorly defined, as it is normally based on scores in tests, examinations, multi-choice quizzes …  None of these assessment-types measure learning. They are proxies for learning and at this time (2018), there is no research including neuroscience efforts that can reliably identify when something is “learned” or, to what that learning can be attributed.  Nor has it been found whether the “learning” is momentary, or long lasting, or useful or not… and those findings would be helpful.

3          VARK does not claim that knowing one’s VARK preference(s) improves academic success. It is farcical to suggest that a student, who completes the VARK Questionnaire and finds out that he/she has a particular VARK preference, will achieve academic success.  For example they could do nothing, and their academic performance might go up or down or they could continue to use the study strategies that they have used before and success might follow or not. It is what the students do, after they know their VARK preference(s) that might make a difference. Knowing your VARK preference does nothing unless the student changes some poor study strategies for some better ones. Then, there might be a difference in academic success.  For example, if a Student X has been using study strategies that are listed by VARK as suitable for a student with a Kinesthetic preference and Student X has a learning preference that is very strongly Visual (according to VARK) it is not likely that there will be any improvement in academic success for Student X unless changes are made.

4         Many researchers have chosen to describe VARK as a “learning style”. For almost a decade VARK has ceased to use that terminology.  VARK is a modality preference and it is one of some 20 plus preferences that affect learning (however that is defined).  If a student has poor motivation then knowing about his/her VARK preference will do very little to improve academic success. Similarly a lack of effort or a lack of persistence or application will also make little difference to academic success no matter what his/her VARK preference is.  If a student uses all the VARK study strategies appropriate for his/her preference and is not motivated, no extra learning or success will occur. Learner’s modality preferences are like any other preference. They might have a preference for biology rather than physics, or a preference for studying alone rather than with others.  Their preferences are subject to change and each preference interacts with others. None of the 20 plus preferences will make an impact merely by identifying it and giving it a name or a score.  Weighing the hog, does not make it fat.

5          There are hypotheses that test whether there is a match between the teachers’ preferences and the students’ preferences. That is fallacious and there is no evidence that a teacher using any particular combination of VARK modalities in teaching will make difference to a student with one or other VARK preference or any combination of preferences.  That concept, called the “matching concept” has been debunked.  It is not what the teacher does that has the biggest impact on learning, but what the student does.  For example, there are studies that have found that the impact of other students on Student X can lead to more academic success than any changes made by a teacher. 

 6          Also needing debunking is the hypothesis that a teacher’s preferences for learning will have an impact on the student(s) success.  Maybe a teacher’s preferences for teaching might have an impact on learning.  Excellent teachers use many and varied strategies regardless of their own preference(s) for learning, and VARK endorses that variety rather than trying to target a learner’s preference(s) often referred to as the matching concept.

 7          There is research that metacognition (learners thinking about and considering how they learn) has benefits for academic success.  Skilled teachers can use VARK to stimulate metacognition among their learners.

8          VARK data is dominated by multimodality with almost 70% of participants choosing two, three or four modalities as their preferences. Yet many researchers attempt to move every participant into one or other of the four VARK modalities. Those learners who have a strong preference for just one modality make up less than a third of the database so moving students who have multimodal preferences into single preference categories is a nonsense process in research.

 9         Some researchers attempt to find a “magic” set of scores that might be linked to academic success or maybe a magic VARK Profile  e.g. ARK (Aural, Read/write and Kinesthetic) as a three-part multi-modality VARK Profile. There are no magic preferences nor combinations of them.  The more positive and more likely hypothesis is that any combination of preferences can be successful for a learner providing he/she uses the study strategies that align with his/her preferences and of course, the student needs to be motivated, persistent, hard working etc. (that is recognising the other 20+ preferences that are factor in learning success).

10          The definitions used by VARK cause some issues with some researchers who believe that the Visual modality means that everything on a screen is a visual and that those learners who  have a Visual preference (according to the VARK questionnaire) will dominate the VARK database. Wrong!!  Visual preferences make up 4% of the VARK database and are not strongly represented in the two, three and four-part multi-modalities.  Much of the learning that appears on a screen involves, talking heads, (Aural), graphs and diagrams Visual) , How-to’s and demonstrations (Kinesthetic) and text (Read/write).  According to VARK data, learners with Visual preferences are not an increasing proportion of the total. 

11        The definition of VARK’s Kinesthetic preference also causes some researchers a problem as they tend to use a simple dictionary definition which is summarised as, “learning by doing” or “action”). That is misleading for VARK.  Learning by doing can involve reading or writing (Read/write) or drawing (Visual) or speaking (Aural). And, a simplistic view that learners with a VARK Kinesthetic preference need to “move around” when they are learning is simplistic and incorrect.  VARK has its own definition of Kinesthetic and that should be understood and used.

 12        The raw scores from the VARK Questionnaires prove difficult for use in many statistical methods because they are not real mathematical numbers.  For example; a student chooses the Visual options for eight of the questions and Read/write options for the other eight questions.   That is a total of 16 and scores of V=8, A=0, R=8, K=0. Another student chooses V and R options for each of eight questions and A for four questions and K for each of the remaining four questions.   That is a total of 24 with scores of V=8, A=4, R=8 and K=4.  The V=8 for the first student and the V=8 for the second student are not statistically comparable. Nor are the Read/write scores of 8 and 8.

 13        VARK scores for individuals need to be interpreted by somebody who is skilled and experienced in using VARK.   The interpreter and the learner still have some work to do when they find out that the learner has VARK scores of 8, 2, 4 and 9,  or 13, 11, 3 and 7.  A researcher or teacher who has not used VARK many times before, or read widely about what it can do, and what it cannot do, is wasting time, effort and resources.

 Requirements when using VARK in research

Paper versions of the questionnaire cannot be used. Participants must be given their VARK preference and related helpsheets after filling in the questionnaire, and the best way for this to be achieved is when using VARK online. To this end, you may need to purchase a VARK Subscription Site or VARK Business Membership.

Once you have read and understood all of the points on this page (if in doubt about anything above, please ask for clarification), and read the copyright information page, permission to use VARK in your research can be requested on the Copyright Permission page.