How We Teach

Assessment of learning styles of undergraduate medical students using theVARK questionnaire and the influence of sex and academic performance

Rathnakar P. Urval,1 Ashwin Kamath,1 Sheetal Ullal,1 Ashok K. Shenoy,1 Nandita Shenoy,2

and Laxminarayana A. Udupa3

1Department of Pharmacology, Kasturba Medical College, Manipal University, Mangalore, Karnataka, India; 2Department ofOral Medicine and Radiology, Manipal College of Dental Sciences, Mangalore, Karnataka, India; and 3Department ofPharmacology, University of West Indies, Barbados

Submitted 19 February 2014; accepted in final form 22 April 2014

Urval RP, Kamath A, Ullal S, Shenoy AK, Shenoy N, Udupa LA.Assessment of learning styles of undergraduate medical students using theVARK questionnaire and the influence of sex and academic performance.Adv Physiol Educ 38: 216–220, 2014; doi:10.1152/advan.00024.2014.—While there are several tools to study learning styles of students, thevisual-aural-read/write-kinesthetic (VARK) questionnaire is a simple,freely available, easy to administer tool that encourages students todescribe their behavior in a manner they can identify with and accept.The aim is to understand the preferred sensory modality (or modali-ties) of students for learning. Teachers can use this knowledge tofacilitate student learning. Moreover, students themselves can use thisknowledge to change their learning habits. Five hundred undergrad-uate students belonging to two consecutive batches in their secondyear of undergraduate medical training were invited to participate inthe exercise. Consenting students (415 students, 83%) were adminis-tered a printed form of version 7.0 of the VARK questionnaire.Besides the questionnaire, we also collected demographic data, aca-demic performance data (marks obtained in 10th and 12th grades andlast university examination), and self-perceived learning style prefer-ences. The majority of students in our study had multiple learningpreferences (68.7%). The predominant sensory modality of learningwas aural (45.5%) and kinesthetic (33.1%). The learning style pref-erence was not influenced by either sex or previous academic perfor-mance. Although we use a combination of teaching methods, there hasnot been an active effort to determine whether these adequatelyaddress the different types of learners. We hope these data will helpus better our course contents and make learning a more fruitfulexperience.

visual-aural-read/write-kinesthetic; learning style; sex; academic per-formance

THAT STUDENTS DIFFER in their preferences of various modalitiesof learning is a fact unlikely to surprise any teacher. Yet, theutilization of this knowledge in a formal way to enhance theteaching process or learning environment has been almostabsent until recently. In medical education, the emphasis oncovering a fixed syllabus, often extensive, in a limited timeperiod with the time-tested method of didactic lecture providedlittle scope for assessment of learning styles and modificationof teaching styles. However, with the growing interest in newerteaching and learning methods worldwide, which is reflected incurrent Medical Council of India directives for undergraduatemedical education, it is an opportune moment to change theage-old teaching styles with due diligence (10). While there areseveral tools to study learning styles of students, the visual,

aural, read/write, and kinesthetic (VARK) questionnaire is asimple, freely available, easy to administer tool that encouragesstudents to describe their behavior in a manner they canidentify with and accept (7). The aim is to understand thepreferred sensory modality (or modalities) of students forlearning. The questionnaire is designed to identify the follow-ing four sensory modalities: visual, aural, read/write, andkinesthetic (8) (hence the acronym VARK). Teachers can usethis knowledge to facilitate student learning. Moreover, stu-dents themselves can use this knowledge to change theirlearning habits.

Kasturba Medical College in Mangalore is a constituentcollege of Manipal University in Karnataka, India. It caters tothe medical education needs of students from all over thecountry. Excellence in medical education is a constant en-deavor rather than a milestone. Newer teaching learning tech-niques are being tried with a gradual shift of emphasis awayfrom passive to active learning. Hence, a formal attempt wasmade to study the learning styles of two consecutive batches ofsecond-year undergraduate medical students. Data from earlierstudies have shown a possible influence of sex and academicperformance on learning styles. However, the results have beenconflicting, with some studies suggesting a sex difference inlearning style preference, whereas others failed to demonstrateany difference (3, 10, 11). Also, differing sensory modalitypreferences for learning among students would likely be as-sumed to result in differing academic performance in a teach-ing environment that is not diverse and hence does not ade-quately address all types of learners. However, studies havefailed to uniformly demonstrate an effect of learning stylepreference on academic performance (1, 2, 4). To the best ofour knowledge, no study has been conducted in India thataddresses the issue of sex and academic performance withregard to learning style preference. Students of Kasturba Med-ical College have diverse geographic and cultural backgrounds.The study of learning style preferences among this studentpopulation using the VARK questionnaire and determining theinfluence of sex and academic performance is likely to add onto our knowledge of these factors that potentially influencelearning style.

METHODS

This study was initiated following the approval of the InstitutionalEthics Committee. It was performed at the Department of Pharma-cology at Kasturba Medical College. Five hundred undergraduatestudents belonging to two consecutive batches in their second-year ofgraduate medical training (after having completed the physiology,biochemistry, and anatomy courses in their first year) were invited to

Address for reprint requests and other correspondence: A. Kamath, Dept. ofPharmacology, Kasturba Medical College, Manipal Univ., Mangalore 575001,Karnataka, India (e-mail: ashwin.kamath@manipal.edu).

Adv Physiol Educ 38: 216–220, 2014;doi:10.1152/advan.00024.2014.

216 1043-4046/14 Copyright © 2014 The American Physiological Society

by 10.220.33.5 on October 11, 2017

http://advan.physiology.org/D

ownloaded from

participate in the exercise. No incentive was given for participation.The purpose of the study was explained to the students, and writteninformed consent was obtained before the VARK questionnaire wasadministered. Besides the questionnaire, we also collected demo-graphic data, self-reported academic performance data (marks ob-tained in 10th and 12th grades of high school and first-year universitymedical examination), and self-perceived learning style preferences.The self-perceived learning style preference was identified along withdemographic and academic performance data. Participants were askedto describe their learning style(s) by choosing from the followingoptions: 1) visual (learning from graphs, charts, flow diagrams, anddemos); 2) aural (learning from speech, lectures, and discussions); 3)reading/writing (learning from reading and writing); and 4) kines-thetic (learning from performing an activity, touch, hearing, smell,taste, and sight). Version 7.0 of the VARK questionnaire in a printedform was used (8). It consisted of 16 questions with 4 options foreach. Each option correlates to a particular sensory modality prefer-ence. Hence, the modality that received the highest marks was thepreferred sensory modality. Since students were free to select morethan one option, multiple modalities of varying combinations could beobtained. The questions describe situations of common occurrence indaily life, thereby relating to an individual’s learning experience.Students were instructed to choose the answer that best explained theirpreference and circle the letter(s) next to it. They could choose morethan one option or leave blank any question that they felt was notapplicable to them. Questionnaires were evaluated on the basis ofpreviously validated scoring instructions and a chart (8). Since each ofthe answers represents a sensory modality preference, the same wascalculated for an individual participant by adding up the responses forall 16 questions. The score for each VARK component for the entirestudy sample was added up and divided by the total number of studyparticipants to obtain mean scores. The entire exercise was completedin 15–20 min, after which the students were asked to return thequestionnaire with demographic data and other details.

Statistical analyses. Sensory modality preferences/VARK modedistributions are expressed as percentages of students in each cate-gory. Scores of individual VARK components are expressed asmeans � SD. Comparison of VARK scores based on sex andacademic performance was done using an independent samples t-testand one-way ANOVA with Tukey’s post hoc test, respectively.Comparison of unimodal sensory modality preference and VARKmode distribution among both sexes was done using a �2-test. SPSS(version 11.5) was used for statistical analyses. Statistical significancewas set at P � 0.05.

RESULTS

Of the 500 undergraduate students invited to participate inthe exercise, 415 students (83%) consented to provide demo-graphic details and answer the VARK questionnaire. Of these,236 students (56.9%) were women.

Overall learning style preferences. One hundred thirty stu-dents (31.3%) showed a unimodal learning style preference.Among the unimodal group, 45.5% of the students wereauditory learners and 33.1% were kinesthetic learners. Pre-ferred sensory modalities among unimodal learners are shownin Fig. 1.

The majority of students (68.7%) preferred more than onesensory modality for learning. The percentages of students whopreferred two, three, and four sensory modalities of learningare shown in Fig. 2. The most common VARK mode distri-bution among students was quadmodal (36.6%) followed byunimodal (31.3%), bimodal (18.1%), and trimodal (14%). Thepreferred combinations of sensory modalities among the bi-modal and trimodal groups are shown in Table 1. Among thequadmodal group, the commonest VARK style based on thescores obtained for individual sensory modalities was kines-thetic-aural-reading/writing-visual learning (12.5%).

Mean VARK scores for individual sensory modalities oflearning are shown in Table 2. The mean score was highest foraural learning (7.11 � 2.72) and lowest for visual learning(4.64 � 2.48).

Sex and VARK. No significant sex difference was seen interms of unimodal or multimodal learning preferences (Fig. 3).There was also no difference in individual sensory modalitypreference among unimodal learners (Fig. 4). Mean VARKscores for individual sensory modalities were also not signifi-cantly different (Table 3).

Academic performance and VARK. Self-reported marks ob-tained in the 10th and 12th grades as well as last universityexamination were combined, and mean marks were dividedinto four categories (�60%, �60% to �70%, �70% to�80%, and �80%). These were compared with individualVARK styles to determine the influence of academic perfor-mance on preference for a particular learning style. Compari-son of the academic performance grade to individual sensorymodality preference did not show any correlation. Table 4

*Total sum is more than 100 due to approximation of decimal places

V5.4

A45.5

R16.2

K33.1

Sensory modality preference among studentswith unimodal learning style (values expressed in

percentage*, n=130)

Visual Aural Read/write Kinesthetic

Fig. 1. Sensory modality preferences among students with a unimodal learningstyle, as assessed by the visual-aural-read/write-kinesthetic (VARK) question-naire. Values are expressed as percentages.

*Total sum is less than 100 due to approximation of decimal places

31.1

18.114

36.6

VARK mode distribution amongundergraduate medical students (values expressed in

percentage*, n=415)

Unimodal Bimodal Trimodal Quadmodal

Fig. 2. VARK mode distribution among undergraduate medical students.Values are expressed as percentages.

How We Teach

217ASSESSMENT OF LEARNING STYLES USING THE VARK QUESTIONNAIRE

Advances in Physiology Education • doi:10.1152/advan.00024.2014 • http://advan.physiology.org

by 10.220.33.5 on October 11, 2017

http://advan.physiology.org/D

ownloaded from

shows mean scores of individual VARK components based onacademic performance.

Perceived sensory modality and VARK. The majority of thestudents considered themselves to be unimodal (44.1%) orbimodal (34.7%) learners (Fig. 5). Twenty-three students didnot indicate their VARK mode preference. Thus, Fig. 5 showsdata from 392 students. Contrary to the assessed VARK style,self-perceived learning styles of students were reading/writing(42.1%) followed by visual (30.1%). Kinesthetic and aurallearning styles constituted 15.8% and 12%, respectively.

DISCUSSION

Our study used the VARK questionnaire to determine thelearning preferences of undergraduate medical students and theinfluence of sex and academic performance on learning styles.The majority of the students in our study had multiple learningpreferences (68.7%). Similar results have been reported byauthors from different geographic locations (2, 9, 11). Thepredominant sensory modality of learning among unimodallearners was aural (45.5%) followed by kinesthetic (33.1%).The same was reflected by mean scores for the various VARKcomponents. Contrary to the assessed learning preference, theself-perceived preference was predominantly unimodal andbimodal (46.7% and 36.7%, respectively). This highlights theself-utility of the VARK questionnaire. Students can under-stand their own learning preferences, which can be contrary totheir own perception. This helps them to actively engage in alearning environment that they would have otherwise per-ceived to be unsuitable.

Teaching in pharmacology in our institution predominantlyconsists of lecture classes using PowerPoint presentations and,to some extent, chalkboard teaching. Practical classes arepredominantly small-group teaching/demonstrations. Thosewho predominantly are kinesthetic learners are least targeted inour current teaching methods. Also important is the fact thatkinesthetic learning was an important component in the ma-jority of the multimodal learners (Table 1). Moreover, merepresentation of a PowerPoint slide might not stimulate thevisual learners unless the slide content is organized in a mannerthat it can be read and understood during the limited time it isprojected even if the student was not listening to the lecture.The same is true for the other sensory modalities. Multimodallearning does help at least to some extent in overcoming someof the deficiencies.

The influence of sex on learning styles is an area of activeresearch. Previous studies have not yielded any uniform resultsin that the learning preference of female students has rangedfrom being predominantly unimodal to predominantly multi-modal (3, 13, 14). In our study, most of the female undergrad-

Table 1. VARK mode distribution among bimodaland trimodal groups

Bimodal Group Trimodal Group

VARKstyle

Number ofstudents

(n � 75 total) PercentageVARKstyle

Number ofstudents

(n � 58 total) Percentage

AK 23 30.7 AKR 6 10.3AR 13 17.3 AKV 5 8.6AV 2 2.7 ARK 12 20.7KA 12 16 ARV 1 1.7KR 5 6.7 AVK 2 3.4RA 3 4 AVR 2 3.4RK 7 9.3 KAR 6 10.3RV 2 2.7 KAV 5 8.6VA 4 5.3 KRA 3 5.2VK 2 2.7 KVA 3 5.2VR 2 2.7 RAK 4 6.9

RKA 1 1.7VAK 3 5.2VKA 1 1.7VKR 1 1.7VRK 3 5.2

V, visual; A, aural; R, reading/writing; K, kinesthetic.

Table 2. Mean scores of individual VARK components

VARK Component Mean Score SD

V 4.64 2.48A 7.11 2.72R 5.59 3.18K 6.94 4.15

p>0.05 for gender comparison of VARK mode distribution, Chi-square test

*Total sum may be less than or more than 100 due to approximation of decimal places

0 20 40 60 80 100

Male

Female

30.7

31.8

21.2

15.7

14.5

13.6

33.5

39

VARK mode distribution among undergraduatemedical students based on gender (values expressed in

percentage*, n=179 for males, n=236 for females)

Unimodal Bimodal Trimodal Quadmodal

Fig. 3. VARK mode distribution among undergraduate medical students basedon sex. Values are expressed as percentages.

p>0.05 for gender comparison of sensory modality preference, Chi-square test

*Total sum may be less than or more than 100 due to approximation of decimal places

0 20 40 60 80 100

Male

Female

5.5

5.3

40

49.3

14.5

17.3

40

28

Sensory modality preference among students withunimodal learning style based on gender (values expressed in

percentage*, n=55 for males, n=75 for females)

Visual Aural Read/write Kinesthetic

Fig. 4. Sensory modality preferences among students with a unimodal learningstyle based on sex. Values are expressed as percentages.

How We Teach

218 ASSESSMENT OF LEARNING STYLES USING THE VARK QUESTIONNAIRE

Advances in Physiology Education • doi:10.1152/advan.00024.2014 • http://advan.physiology.org

by 10.220.33.5 on October 11, 2017

http://advan.physiology.org/D

ownloaded from

uate medical students (68.3%) preferred more than one modal-ity of learning. Among those with unimodal preferences, fe-male students were predominantly aural (49.3%), whereas anequal number of male students preferred aural and kinesthetic(40%). Unlike a previous study (13) that showed that amongstudents with multimodal learning preferences female studentshad a more diverse combination of sensory modalities, ourstudy showed diverse sensory modality combinations in bothsexes. Hence, in our study sample, sex did not significantlyinfluence the learning style preference. Considering the differ-ing results from various studies, no generalizations can bemade regarding the influence of sex.

Our study also assessed the influence of academic perfor-mance on learning style preference. No correlation was foundbetween the academic performance of students and their pre-ferred VARK style. Few previous studies have addressed thesame issue. No definite conclusions can be drawn from thesestudies. A study among nurses showed a correlation betweengood performance and multimodal learning (1). However, asimilar study among dental students showed no statisticalassociation (2). Two studies among physiology studentsshowed a relatively poor performance by students who pre-ferred the kinesthetic modality (3, 4). Similar to the influenceof sex, no generalized conclusions could be drawn with regardto the influence of academic performance. The VARK ques-tionnaire does not provide for a complete assessment of learn-ing style, nor was it intended to. Learning is influenced bymultiple components, such as psychological, social, physical,and environmental elements (5). For example, nurses andlawyers require different skill sets, which would influence theirlearning style, including adopting or getting used to a learning

style that one may not be comfortable with at the beginning(13). Similarly, translating the effects of an improved teachingmethod that caters to all types of learners into improvedacademic performance or learning has been difficult to dem-onstrate considering the lack of a definition for learning andmethods to measure it (6).

Our study has limitations. The academic performance datawas based on self-reported marks obtained in the variousexaminations. This could have led to recall bias or inaccuratereporting. Our study does not address whether altering theteaching methods according to student learning styles improvesacademic performance. It only tells us that differing pastacademic performance did not significantly influence the learn-ing style preference of students. Although our study showed noinfluence of sex on learning style preferences, a larger studysample might have shown a statistically significant differenceconsidering that our study sample showed more kinestheticlearners among men, which was not statistically significant(40% vs. 28% for men vs. women, respectively). While col-lecting data on self-perceived learning style preferences, stu-dents were allowed to select multiple preferences. Hence, wewere unable to determine the self-perceived dominant learningmodality.

In conclusion, our study was an attempt to describe thelearning styles of undergraduate medical students in our insti-tution. Based on the questions raised by previous studies, wealso studied the influence of sex and previous academic per-formance on learning style preference. Most of the studentswere multimodal learners, which is good from both a teachingas well as learning perspective. Aural and kinesthetic were thepreferred sensory modalities of learning. Neither sex nor pre-vious academic performance had any relation to learning style

Table 3. Mean scores of individual VARK components basedon sex

VARK Component Mean Score* SD

VMen 4.59 2.41Women 4.68 2.54

AMen 7.17 2.92Women 7.06 2.57

RMen 5.40 3.96Women 5.72 2.43

KMen 7.21 5.69Women 6.74 2.39

*P � 0.05 for sex comparison in all groups by an independent samplest-test.

Table 4. Mean scores of individual VARK components based on academic performance

Academic Performance, %

VARK Component*

V A R K

�60 5.33 � 3.21 4.67 � 3.51 5.00 � 2.65 4.00 � 3.00�60 to �70 5.00 � 2.58 6.59 � 2.73 5.00 � 2.61 6.08 � 2.13�70 to �80 4.33 � 2.51 7.26 � 2.72 5.61 � 2.70 7.47 � 6.25�80 4.78 � 2.45 7.10 � 2.72 5.64 � 3.49 6.76 � 2.32

Values are means � SD. *P � 0.05 for comparison of sensory modality of learning and academic performance by one-way ANOVA with a Tukey’s post hoctest.

46.7

36.7

115.6

Self perceived VARK mode distributionamong undergraduate medical students (values

expressed in percentage, n=392)

Unimodal Bimodal Trimodal Quadmodal

Fig. 5. Self-perceived VARK mode distribution among undergraduate medicalstudents. Values are expressed as percentages.

How We Teach

219ASSESSMENT OF LEARNING STYLES USING THE VARK QUESTIONNAIRE

Advances in Physiology Education • doi:10.1152/advan.00024.2014 • http://advan.physiology.org

by 10.220.33.5 on October 11, 2017

http://advan.physiology.org/D

ownloaded from

preference. Although we use a combination of teaching meth-ods, there has not been an active effort to determine whetherthese adequately address the different types of learners. Wehope these data will help us better our course contents andmake learning a more fruitful experience.

ACKNOWLEDGMENTS

The authors thank Neil D. Fleming for permission to use the VARK question-naire [copyright version 7.0 (2006) held by Neil D. Fleming, Christchurch, NewZealand, and Charles C. Bonwell, Green Mountain Falls, CO 80819].

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

Author contributions: R.P.U., S.U., A.K.S., N.S., and L.A.U. conceptionand design of research; R.P.U. and S.U. performed experiments; R.P.U. andA.K. analyzed data; R.P.U. and A.K. interpreted results of experiments;R.P.U., A.K., S.U., A.K.S., N.S., and L.A.U. edited and revised manuscript;R.P.U., S.U., A.K.S., N.S., and L.A.U. approved final version of manuscript;A.K. prepared figures; A.K. drafted manuscript.

REFERENCES

1. Alkhasawneh IM, Mrayyan MT, Docherty C, Alashram S, YousefHY. Problem-based learning (PBL): assessing students’ learning prefer-ences using VARK. Nurs Educ Today 28: 572–579, 2008.

2. Baykan Z, Nacar M. Learning styles of first-year medical studentsattending Erciyes University in Kayseri, Turkey. Adv Physiol Educ 31:158–160, 2007.

3. Dobson JL. A comparison between learning style preferences and sex,status and course performance. Adv Physiol Educ 34: 197–204, 2010.

4. Dobson JL. Learning style preferences and course performance in anundergraduate physiology class. Adv Physiol Educ 33: 308–314, 2009.

5. Felder RM, Brent R. Understanding student differences. J Eng Educ 94:57–72, 2005.

6. Fleming ND. The Case Against Learning Styles: “There Is No Evi-dence . . .” (online). http://www.vark-learn.com/documents/The%20Case%20Against%20Learning%20Styles.pdf (1 May 2014).

7. Fleming ND, Mills C. Not Another Inventory, Rather a Catalyst forReflection (online). http://digitalcommons.unl.edu/cgi/viewcontent.cgi?article�1245&context�podimproveacad (1 May 2014).

8. Fleming ND. VARK: a Guide to Learning Styles (online). http://www.vark-learn.com (1 May 2014).

9. Lujan H, DiCarlo S. First-year medical students prefer multiple learningstyles. Adv Physiol Educ 30: 13–16, 2006.

10. Medical Council of India. Medical Council of India Regulations onGraduate Medical Education, 1997 (online). http://www.mciindia.org/Rules-and-Regulation/GME_REGULATIONS.pdf (1 May 2014).

11. Nuzhat A, Salem RO, Quadri MSA, Al-Hamdan N. Learning stylepreferences of medical students: a single-institute experience from SaudiArabia. Int J Med Educ 2: 70–73, 2011.

12. Slater JA, Lujan HL, DiCarlo SE. Does gender influence learning stylepreferences of first-year medical students? Adv Physiol Educ 31: 336–342,2007.

13. Academic Learning Centre, University of Manitoba. Learning Styles(online). https://umanitoba.ca/student/academiclearning/media/Learning_Styles_NEW.pdf [1 May 2014].

14. Wehrwein EA, Lujan HL, DiCarlo SE. Gender differences in learningstyle preferences among undergraduate physiology students. Adv PhysiolEduc 31: 153–157, 2007.

How We Teach

220 ASSESSMENT OF LEARNING STYLES USING THE VARK QUESTIONNAIRE

Advances in Physiology Education • doi:10.1152/advan.00024.2014 • http://advan.physiology.org

by 10.220.33.5 on October 11, 2017

http://advan.physiology.org/D

ownloaded from