Neil D Fleming
October 2002
This is a set of descriptive statistics about the VARK database. Before we analyse the results from the VARK database it is necessary to examine the shape and structure of the questionnaire so that the correct statistical techniques can be used on it.
WHAT DOES VARK INDICATE?
VARK is not a fully-fledged learning style. The word learning style is now used loosely to describe almost any attribute or characteristic of learning but technically the term refers to all the components that might affect a person’s ability to learn. Some inventories report on a number of components in a style (motivation, surface-deep approaches to learning, social and physical and environmental considerations) and some personality inventories have learning characteristics as a part of their descriptions.
VARK deals with only one dimension of the complex amalgam of preferences that make up a learning style. The VARK questions and results focus on the ways in which people like information to come to them and the ways in which they like to deliver their communication. The questions are based on situations where there are choices about how that communication might take place. It is important to say what VARK is not, so that other components are not perceived as being a part of it. VARK has little to say about personality, motivation, social preferences, physical environments, intraversion-extraversion… The choice to limit VARK to modal preferences was made because that is where its designer had most success in assisting with learning. Changing the other dimensions affected learning, of course, but it was the modal preferences that had the most direct application for helping both students and faculty learn.
THE RATIONALE FOR MULTIPLE CHOICES.
Multimodality was the expectation in the questionnaire design. The modal preferences of people are seldom singular. In the majority of cases people will have preferences for a number of modes and they will use strategies associated with their preferences depending on the context or situation. For example they may choose a Read/write response because the situation is biased towards it. Intuitively this makes sense as we seldom act on the basis of input or output from only one mode. For that reason, multimodality (bi-, tri- or quad-) is likely to be the “normal” condition and single-preferences are likely to be less common. Those who have a mild, strong or very strong preference for one mode are still multimodal – it is just that one of their preferences is a little stronger than the others. For example a person with VARK scores of 6 3 3 3 is said to have a single preference for V but is, in fact, still multimodal, though not categorised as such by the VARK algorithm. Some modes, notably K, is itself an amalgam of senses and could be said to be multimodal in the broader sense of that word. An alternative hypothesis, that most people have a uni-modal approach to communication has low probability though it is not impossible.
If multimodality is the expectation in life situations we should allow for it in the structure of the VARK questionnaire and that is why respondents can choose more than one answer to each question. But clearly if everyone chose every answer for every question then VARK would provide few insights into strategies for learning. Allowing for multiple choice, however, reduces the discrimination of VARK. So on one hand we say that multimodality is the norm but on the other hand we are really interested in the relative strengths of particular modes within individuals. It is the ability of VARK to allow multiple choices, yet point out a person’s established preferences in their profile, that is its strength.
SINGLE PREFERENCES
If the database indicated that the respondents choices were distributed evenly across all options then it is likely that the questionnaire would provide less discriminated information for its respondents – most would be VARK. The options to each question are designed so that those with a particularly strong preference will choose the response that matches their preference even when the situation in the question stem is biased towards another mode. That is how VARK discriminates and for that reason the proportion of respondents choosing each option in a question is unlikely to be close to 25% for the four-option questions or 33% for the three-option questions. It is more likely that one or sometimes two options in each question will be very attractive and that only those with a strong preference will choose alternative answers containing their modal preference. We hypothesise that those who have a single-preference continue to choose the “weakest” option despite the attraction of the dominant option (see later). So an uneven distribution across the options is expected. Table One shows this feature in the proportions (percentages) for each question taken from the June-September 2002 database (n=31243).
For ten of the 13 questions there is a dominant or popular choice above an arbitrary benchmark of 37%. For ten questions there is a weak choice (sometimes two) below 15%. V appears as the lowest choice in six questions, A in four, R in one and K in two. Conversely, V is the main choice for three questions, A for two, R for three and K for five questions.
TABLE ONE: Percentages choosing each option
|
Proportion who chose this option
|
|||||||
|
Question
|
V
|
A
|
R
|
K
|
Total
|
Most popular option
|
Least popular Option
|
|
1
|
33.8%
|
28.0%
|
28.2%
|
10.1%
|
100%
|
V
|
K
|
|
2
|
11.1%
|
10.8%
|
43.8%
|
34.3%
|
100%
|
R
|
A
|
|
3
|
5.8%
|
48.6%
|
26.5%
|
19.1%
|
100%
|
A
|
V
|
|
4
|
36.2%
|
30.5%
|
33.3%
|
100%
|
V
|
R
|
|
|
5
|
11.4%
|
15.9%
|
26.3%
|
46.3%
|
100%
|
K
|
V
|
|
6
|
12.8%
|
12.7%
|
36.5%
|
38.0%
|
100%
|
K
|
A
|
|
7
|
14.5%
|
21.0%
|
15.5%
|
49.0%
|
100%
|
K
|
V
|
|
8
|
26.7%
|
43.8%
|
29.4%
|
100%
|
A
|
V
|
|
|
9
|
17.2%
|
27.3%
|
55.5%
|
100%
|
K
|
A
|
|
|
10
|
29.3%
|
29.7%
|
30.8%
|
10.2%
|
100%
|
R
|
K
|
|
11
|
11.9%
|
30.0%
|
42.7%
|
15.4%
|
100%
|
R
|
V
|
|
12
|
69.2%
|
9.1%
|
21.7%
|
100%
|
V
|
A
|
|
|
13
|
14.4%
|
23.8%
|
24.0%
|
37.8%
|
100%
|
K
|
V
|
Note: The table was calculated using all the choices for all respondents for each question. Many respondents chose more than one option.
NUMBER OF OPTIONS CHOSEN
For most questions, (see Table Two) over three-quarters of the respondents chose only one option. Ninety five percent chose one or two options and the percentage that chose all four options was greater than 1% for only four of the nine four-option questions – 5, 6, 7 and 13. The percentage that chose all the three-options questions (4, 8, 9 and 12) was 1%, 4%, 3% and 3% respectively. The fact that single responses were popular may be related to the custom for most questionnaires to allow only one answer per question or to the idea that there are “right answers’ in the VARK questionnaire.
TABLE TWO : How many options did each respondent choose for each question? (%)
|
Question
|
One option
|
Two options
|
Three options
|
Four options
|
No choice
|
Total
|
|
1
|
71%
|
25%
|
4%
|
0.5%
|
0.2%
|
100%
|
|
2
|
79%
|
18%
|
3%
|
0.9%
|
0.2%
|
100%
|
|
3
|
78%
|
17%
|
2%
|
0.7%
|
1.0%
|
100%
|
|
4
|
88%
|
10%
|
1%
|
N.A.
|
0.9%
|
100%
|
|
5
|
72%
|
20%
|
5%
|
2.5%
|
1.3%
|
100%
|
|
6
|
70%
|
24%
|
5%
|
1.0%
|
0.3%
|
100%
|
|
7
|
65%
|
26%
|
6%
|
2.5%
|
0.2%
|
100%
|
|
8
|
79%
|
17%
|
4%
|
N.A.
|
0.4%
|
100%
|
|
9
|
76%
|
21%
|
3%
|
N.A.
|
0.4%
|
100%
|
|
10
|
72%
|
24%
|
4%
|
0.5%
|
0.2%
|
100%
|
|
11
|
74%
|
20%
|
4%
|
0.8%
|
1.1%
|
100%
|
|
12
|
84%
|
12%
|
3%
|
N.A.
|
0.6%
|
100%
|
|
13
|
64%
|
24%
|
7%
|
3.8%
|
0.5%
|
100%
|
|
Mean
|
75%
|
20%
|
4%
|
1%
|
0.6%
|
N.A. = not applicable to this question.
THE VARK PREFERENCE SETS – PROFILES
VARK allows for the condition of multimodality but attempts to find, within that obvious multimodality, some strengths for particular preferences. This is similar to saying that everyone has a V, A, R and K profile but some have particular sets of preferences within that. The 23 possible profiles of preferences that VARK identifies are as follows:
| VARK Profiles | Number | |
| Single preferences | V Mild, Strong and Very strong. A Mild, Strong and Very strong. R Mild, Strong and Very strong. K Mild, Strong and Very strong. |
12 |
| Bi-modal preferences | VA VR VK AR AK RK | 6 |
| Tri-modal | VAR VAK ARK VRK | 4 |
| All four modes | VARK | 1 |
| Total number of possible profiles | 23 |
TABLE THREE: Profiles: Database Distribution of Types
n=31243
|
Profile
|
%
|
Mild
|
Strong
|
Very Strong
|
Subtotals
|
%
|
|
V
|
3.2
|
1.8
|
0.9
|
0.5
|
||
|
A
|
5.3
|
2.7
|
1.7
|
0.9
|
||
|
R
|
15.6
|
5.9
|
5.0
|
4.7
|
||
|
K
|
17.6
|
7.4
|
5.7
|
4.5
|
||
|
Single preference
|
41.7
|
|||||
|
VA
|
0.4
|
|||||
|
VR
|
1.2
|
|||||
|
VK
|
3.1
|
|||||
|
AR
|
1.9
|
|||||
|
AK
|
3.4
|
|||||
|
RK
|
4.6
|
|||||
|
Bi modal
|
14.6
|
|||||
|
VAR
|
0.8
|
|||||
|
VAK
|
2.7
|
|||||
|
ARK
|
5.0
|
|||||
|
VRK
|
4.3
|
|||||
|
Tri modal
|
12.8
|
|||||
|
VARK
|
30.8
|
30.8
|
||||
|
Total
|
100%
|
100%
|
||||
This means for example that there is some V in several profiles – in the three single V preferences Mild, Strong and Very Strong, and in VA, VR and VK, VAR, VAK, VRK and in VARK. Each mode is therefore represented in 10 profiles, seven of which are overlapping with other modes.
SO WHAT IS NORMAL?
The database samples mainly those who are students and teachers so it is not representative of the total population. Like most databases it also contains duplicates from respondents who visit the site more than once. In the absence of a distinctive distribution what does it mean if R and K are chosen more often? It could be argued that the proportions above indicate biases towards R and K or, that the questionnaire measures what it measures, and that that, is all it does. Is that why there is a trend in the results towards lower proportions for V and A? But this is a circular argument. In questionnaires where only one option can be selected there is a balancing effect. Choosing one option precludes another so if one set of choices is popular then, by definition there will be other less popular choices. If the questions and options were rewritten to balance the proportions we are merely reflecting an hypothesis that modal preferences are balanced within our society. That is a contestable hypothesis and the statistics don’t help us decide, as they are a result rather than a cause.
The content validity of VARK is the best source for resolving this argument. The strength of VARK is that its questions and options are drawn from real life situations and that people identify with the results that they receive. In that lies VARK’s strength. If people found the questionnaire gave them results outside their own perceptions or the perceptions of those close to them, then that would be a reason to re-examine the questionnaire. If we really wanted to balance the results for V, A, R and K we should search/create a number of additional questions so that V and A are the dominant options more often?
As a consequence there is no distinctive distribution of VARK scores – no typical VARK profile for the general population. What we can describe is the most common final scores for each mode (V, A, R and K) which are 3, 3, 4 and 5.
V, A, R and K PROPORTIONS
The database has 31243 entries but with some respondents having a multimodal profile, the total number of modes represented in the statistics is more than twice that number (72662). If we total all the respondents who had some V, some A some R and some K the results are as in Table Four below . The following calculation is used: -a person with a VA profile is counted as a contributor to both V and A. A person with a VAR profile contributes three modes to the count. Table Four shows the relative proportions of the four modes in the sample.
TABLE FOUR: V, A, R and K proportions (Method One).
| V | A | R | K | ||
| % | 20% | 22% | 28% | 31% | 100% |
| No. | 14536 | 15710 | 20080 | 22336 | 72662 |
An alternative method of counting is to limit each respondent’s contribution to the count as a single unit by apportioning the modes, ascribing 1.0 for all single modes, 0.5 for each mode where there are two, 0.333 for each mode where there are three and 0.25 where there are four. This is used for Table Five below.
| V | A | R | K | ||
| % | 16% | 19% | 31% | 35% | 100% |
| No. | 4964 | 5841 | 9545 | 10894 | 31243 |
Note that the total (31243) is the total number of respondents in the database.
TABLE SIX: Groups and the proportions of V, A, R and K.
| V | A | R | K | n= | |
| Total | 16% | 19% | 31% | 35% | 31243 |
| Males % | 16% | 18% | 29% | 38% | 10471 |
| Females % | 16% | 19% | 32% | 33% | 20173 |
| Students | 16% | 20% | 29% | 36% | 22380 |
| Teachers | 17% | 16% | 35% | 32% | 5673 |
| Applied science | 16% | 18% | 30% | 36% | 2292 |
| Business | 15% | 20% | 31% | 35% | 4342 |
| Education | 16% | 19% | 29% | 35% | 4971 |
| Humanities | 16% | 19% | 33% | 32% | 2097 |
| Other | 16% | 19% | 29% | 37% | 6882 |
| Science | 17% | 18% | 30% | 35% | 4564 |
| Social science | 15% | 21% | 32% | 33% | 2261 |
| Employee | 15% | 18% | 34% | 33% | 5268 |
| Employer | 16% | 21% | 32% | 31% | 350 |
| Other | 15% | 19% | 31% | 35% | 1845 |
| Unwaged | 15% | 20% | 34% | 31% | 365 |
| Education | 16% | 19% | 30% | 35% | 26194 |
| Not in education | 16% | 18% | 35% | 31% | 3076 |
| Pre-university | 16% | 19% | 27% | 38% | 6680 |
| University | 16% | 19% | 31% | 34% | 20903 |
A phenomenon that has been present in VARK since its design is the Alps/Staircase nature of the teacher/student proportions for each of V, A, R and K. The student graph rises from V to K (Staircase). The faculty graph drops from V to A then rises to R and drops again to K (Alps).
The interesting statistics in Table six are:
- The proportion of females who complete the questionnaire compared with males. Why are there almost twice as many women?
- The Alps/Staircase shape of the students and teachergraphs.
- The remarkable symmetry in the results for the subsets. Few groups seem to vary significantly from the total database proportions of 16%, 19%, 31% and 35%. Those that do are often small subsets – waged (n=3650 employer (n=350) or those who could have used the VARK for Younger questionnaire that has questions more appropriate for those aged 12-18. (n=-6680).
- Most of the variation is in the R and K proportions.
THE DISTRIBUTION OF V, A, R AND K SCORES
The VARK website algorithm calculates each respondent’s profile based on their V, A, R and K scores from the questionnaire. The scores vary from zero (0) to 12. The frequency of each V, A, R and K score for all respondents is shown in the graphs below.
The shape of the kinesthetic graph is different from the other three that are remarkably similar (skew). The reasons for that are not known but the effect is to lift the representation of K in respondents’ profiles.
The distribution of those who had a zero score (voids) for a particular mode is shown in Table Seven as is the number of respondents who chose all twelve options for a particular mode. Note that 31 respondents chose all 48 options in the questionnaire!
TABLE SEVEN: Voids and All Options (n=31243)
| Mode | Number who chose no options for this mode. | % of total respondents | Number who chose all options for this mode |
| V | 633 | 2.0% | 39 |
| A | 1563 | 5.0% | 66 |
| R | 798 | 2.6% | 134 |
| K | 237 | 0.8% | 53 |
NUMBER OF OPTIONS CHOSEN
VARK has 13 questions with 48 options spread across those questions. Each of the four modes (V, A, R and K) can be selected 12 times. Because each respondent may choose more than one answer for each of the questions the possible number of answers for any single respondent is 48 (12 V, 12 A, 12 R and 12 K). For a valid entry in the database the minimum number of questions attempted has been set at 10. Just over forty percent (40.6%) chose thirteen options (Note: not necessarily one per question) and 85% of respondents chose between 10 and 20 options.
THE APPROPRIATENESS OF THE OPTIONS.
It might be helpful to know which options were “working” and which were not. One possibility for VARK was to design options so that each would attract significant numbers of respondents. This would have led to almost equal proportions of respondents opting for each choice – for each question there would be equal numbers of respondents choosing each option. But because VARK allows multiple answers to each question and because there is no right or correct answer, the proportions are variable. To test this we looked at the least preferred option in each question and analysed who was choosing it. The hypothesis was that as long as it was being chosen by a number of respondents who had that preference then it was assisting with the overall discrimination.
Who chose the weakest options?
If the weakest option was still being selected by those who had some preference for the mode represented in that option then one could be confident that the questionnaire and its options were working appropriately. The alternative hypothesis that those who had no preference for the mode were choosing that option would indicate that the option was wrongly worded or poorly selected. Table Eight indicates the percentage of respondents who had a single VARK preference that matched the weakest option for each question. The percentage who had the mode of the weakest option (V, A, R or K) somewhere in their profile is also shown. The hypothesis is that a substantial proportion of those with a single preference for the weakest option will have chosen it and that a number of those with that mode as part of their profile will also have chosen it. So where V is the weakest option in a question we would expect that a substantial number of those with V as part of their profile (VA, VR, VK, VAR, VRK, VAK, and VARK) and those with V as their single preferred mode (Mild, Strong or Very strong) will have chosen it.
In Table Eight below the third column shows that 27% of those with a single preference for K chose K for this question. The fourth column shows, for example, that 43% of all those who chose K in question 1 had K as part of their profile (single or multimodal). There is wide variation in these data. Note that questions 4, 8, 9 and 12 have only three options so one would expect their percentages for the weakest option to be higher. That is not the case for question 12.
TABLE EIGHT: VARK profiles of those selecting the weakest options.
N=31243
| Question Number | Weakest option | Percentage of those with a single preference for the mode in column 2 who chose that option in this question. | Percentage of those who chose the weakest option who had the mode in Column 2 as part of their profile. |
| 1 | K | 27 | 43 |
| 2 | A | 27 | 28 |
| 3 | V | 23 | 28 |
| 4 | R | 57* | 35* |
| 5 | V | 36 | 26 |
| 6 | A | 43 | 29 |
| 7 | V | 48 | 27 |
| 8 | V | 66* | 25* |
| 9 | A | 54* | 29* |
| 10 | K | 25 | 37 |
| 11 | V | 41 | 27 |
| 12 | A | 21* | 28* |
| 13 | V | 51 | 27 |
* Three options in these questions.
SOME INTERESTING PROFILES
Those with voids (zero scores for a mode) have interesting profiles as do those who choose all 12 of a particular mode. Here is a sampling of those from the database.
Those with Double Voids (sample only)
No V or A
| V | A | R | K | Profile | Sex | Education/ Not education | Student/ teacher | University/ Pre-university | Discipline | Employed? |
| 0 | 0 | 9 | 5 | very strong R | female | education | student | university | education | |
| 0 | 0 | 10 | 3 | very strong R | female | education | student | university | education | employee |
| 0 | 0 | 7 | 6 | RK | male | education | student | university | applied science | |
| 0 | 0 | 7 | 6 | RK | female | education | teacher | university | other | employee |
| 0 | 0 | 7 | 6 | RK | male | student | university | other | ||
| 0 | 0 | 3 | 10 | very strong R | female | education | student | university | other | |
| 0 | 0 | 3 | 10 | very strong R | female | Not in education | other | |||
| 0 | 0 | 6 | 7 | RK | male | Not in education | teacher | employee | ||
| 0 | 0 | 8 | 5 | strong R | male | education | teacher | university | science | |
| 0 | 0 | 9 | 4 | strong R | male | education | teacher | university | social science | |
| 0 | 0 | 7 | 6 | RK | female | education | student | university | science | |
| 0 | 0 | 7 | 6 | RK | male | Not in education | student | university | applied science | employee |
| 0 | 0 | 9 | 4 | very strong R | female | education | student | university | other | |
| 0 | 0 | 6 | 7 | RK | female | Not in education | employee | |||
| 0 | 0 | 5 | 11 | very strong R | female | education | student | university | education |
No A or R
| V | A | R | K | Profile | Sex | Education/ Not education | Student/ teacher | University/ Pre-university | Discipline | Employed? |
| 11 | 0 | 0 | 2 | very strong V | female | education | student | university | social science | |
| 10 | 0 | 0 | 3 | very strong V | male | education | student | university | science | |
| 9 | 0 | 0 | 4 | very strong V | female | education | student | university | science | |
| 11 | 0 | 0 | 4 | very strong V | female | education | student | Pre university | science | |
| 8 | 0 | 0 | 5 | strong V | female | education | student | university | social science | |
| 8 | 0 | 0 | 5 | strong V | male | education | student | Pre university | other | |
| 8 | 0 | 0 | 5 | strong V | female | education | teacher | Pre university | other | |
| 10 | 0 | 0 | 5 | very strong V | female | education | teacher | university | education | |
| 5 | 0 | 0 | 6 | VK | male | education | student | university | applied science | |
| 8 | 0 | 0 | 6 | mild V | female | education | student | university | science | |
| 8 | 0 | 0 | 6 | mild V | female | education | teacher | Pre university | education | employee |
| 6 | 0 | 0 | 7 | VK | female | education | teacher | Pre university | humanities | |
| 6 | 0 | 0 | 7 | VK | female | education | teacher | Pre university | education | employee |
| 6 | 0 | 0 | 7 | VK | female | education | student | university | science | |
| 5 | 0 | 0 | 8 | strong K | female | education | teacher | university | science | |
| 5 | 0 | 0 | 8 | strong K | female | education | student | university | social science | |
| 5 | 0 | 0 | 8 | strong K | male | education | student | university | applied science | |
| 5 | 0 | 0 | 8 | strong K | male | education | student | university | social science | |
| 4 | 0 | 0 | 9 | very strong K | female | education | teacher | Pre university | other | other |
| 4 | 0 | 0 | 9 | strong K | female | education | student | Pre university | other | |
| 4 | 0 | 0 | 9 | very strong K | female | education | student | university | science | |
| 4 | 0 | 0 | 9 | very strong K | male | education | student | Pre university | business | |
| 4 | 0 | 0 | 9 | very strong K | female | education | teacher | university | education | |
| 6 | 0 | 0 | 9 | strong K | female | education | student | university | other | |
| 3 | 0 | 0 | 10 | very strong K | female | education | student | Pre university | education | |
| 3 | 0 | 0 | 10 | very strong K | male | education | student | university | education | |
| 4 | 0 | 0 | 10 | very strong K | female | education | student | university | other | |
| 2 | 0 | 0 | 11 | very strong K | male | education | student | Pre university | science | |
| 3 | 0 | 0 | 11 | very strong K | female | education | student | university | humanities | employee |
Some of the 31 who chose all the options in all the questions!
| V | A | R | K | Profile | Sex | Education/ Not education | Student/ teacher | University/ Pre-university | Discipline | Employed? |
| 12 | 12 | 12 | 12 | VARK | male | education | student | pre university | science | |
| 12 | 12 | 12 | 12 | VARK | male | education | student | university | social science | |
| 12 | 12 | 12 | 12 | VARK | male | education | student | university | applied science | employer |
| 12 | 12 | 12 | 12 | VARK | male | education | student | pre university | science | |
| 12 | 12 | 12 | 12 | VARK | male | education | student | pre university | business | |
| 12 | 12 | 12 | 12 | VARK | female | education | student | university | business | employer |
| 12 | 12 | 12 | 12 | VARK | female | not in education | teacher | pre university | business | unwaged |
| 12 | 12 | 12 | 12 | VARK | education | student | university | other | employer | |
| 12 | 12 | 12 | 12 | VARK | male | education | student | university | business | |
| 12 | 12 | 12 | 12 | VARK | female | education | student | university | other | unwaged |
| 12 | 12 | 12 | 12 | VARK | male | education | student | pre university | science | |
| 12 | 12 | 12 | 12 | VARK | education | teacher | university | other | ||
| 12 | 12 | 12 | 12 | VARK | male | education | student | pre university | business | |
| 12 | 12 | 12 | 12 | VARK | male | education | student | university | business | other |
| 12 | 12 | 12 | 12 | VARK | male | education | student | university | applied science | |
| 12 | 12 | 12 | 12 | VARK | male | education | student | pre university | other | |
| 12 | 12 | 12 | 12 | VARK | male | not in education | student | pre university | science | other |
| 12 | 12 | 12 | 12 | VARK | female | education | student | pre university | science | |
| 12 | 12 | 12 | 12 | VARK | male | not in education | student | employer | ||
| 12 | 12 | 12 | 12 | VARK | male | education | student | pre university | other | |
| 12 | 12 | 12 | 12 | VARK | female | education | teacher | university | applied science | unwaged |
| 12 | 12 | 12 | 12 | VARK | female | education | teacher | university | other | unwaged |
| 12 | 12 | 12 | 12 | VARK | female | education | student | university | other |
