My apologies for the delay between posts- I have been reading through quite a few learning style papers out there and had some trouble deciding where to go with my next post. The vast majority of my reading focused on the topic of Neuro-Linguistic Programming, which will be the focus of a post in the near future.
For the current post, I decided to review a paper published last year by John Dobson from the University of Florida. Dobson published a paper looking at learning style preferences in an exercise physiology class. The article is written with the assumption that learning styles exist; so, the purpose of Dobson’s paper wasn’t to examine the existence of learning styles. His purpose appears to be to examine the proportion of learner type amongst groups of students in an exercise physiology class. Specifically, he wanted to know whether graduate students and undergraduate students have different learning style preferences. He also wanted to know whether a difference in preferences existed across genders and, at the end of the study, he examines whether course performance had anything to do with learning style preference.
In the study, Dobson compared perceived modality preferences and assessed modality preferences across his domains of interest. Basically, he asked students what they thought was their modality preference for learning (Visual, Auditory, Reading, or Kinesthetic). He called this their perceived learning style preference. Then he gave them the VARK questionnaire to assess their learning style preference. He called this their assessed learning style preference. Some similarities were found as well as some interesting differences.
The study found that 59% of the time the perceived modality preference and the assessed modality preference matched. The authors claim that this matches the numbers that Neil Fleming has found (see previous posts for discussion of Fleming’s work). A closer look at the data reveals that this number is not very impressive. 64 students completed the study. According to the data, 38 students (59%) had a match between their perceived modality preference and the assessed modality preference. However, in this study, many students (24) were classified as VARK learners by the assessment (which means that they have no real preference, but instead are happy receiving information in either visual, auditory, reading, or kinesthetic modality). It is not very impressive that these particular 24 students matched their VARK score and their perceived preference; given that they have no specific preference, they could have answered ANYTHING on the perceived preference portion of the experiment and Dobson would have concluded that their perceived and assessed learning style preferences “matched.” Again, that means 24 of the 38 that “match” would have matched regardless of what they had listed for their perceived preference!
So, we can get a more accurate gauge about how the perceived preferences and the assessed (ie VARK) preferences matched by excluding that group of 24 students. If we do this, there are 40 students that had a specific preference of one or more modality (ie. all students that were not classified as a VARK learner). Only 14 of those 40 matched with their perceived preferences. That is only a 35% concordance rate for the VARK questionnaire matching perceived preferences, which is quite low and may question the validity of the assessment tool. In another 14 of these 40 students, VARK found a minor preference for their perceived learning style preference. That leaves 12 students (of the 40) where VARK didn’t even list off their perceived learning style preference at all. So these data from this study question the validity of the VARK assessment. How accurate is VARK if it doesn’t match perceived preferences in a majority of cases?
The study wanted to examine whether status, gender, and class performance (grade) varied across learning style preference. For status, the study found no difference between preferences of graduate and undergraduate students (either perceived preferences or assessed preferences).
For gender, the study found no significant differences between preferences of males and females. However, it should be noted that the test statistics were very close to significance for the gender domain. The author claims that this may indicate that a difference really does exist, but due to a small sample (only 40 women and 24 men) they did not achieve significance. A closer look into the data suggests that there might be more going on here than meets the eye. As previously discussed, there were large divergences between assessed preferences and perceived preferences. This was when comparing males and females as well. The perceived preference data shows different patterns for males and females: R preference: 35% female, 17% male; V preference: 25% female, 54% male; A preference: 22% female, 8% male and K preference: 18% female, 21% male. Notice the largest differences between male and female here are in the visual modality and in the reading modality, with the kinesthetic modality having nearly equivalent proportions of males and females.
Assessed preferences show that largest gender difference between those that are K learners (13% male, 5% female), those that are A learners (7% female, 0% male), and many of the combinations of modality preferences (ie the VARK will classify people as a VK (visual and kinesthetic) or a VA (visual and auditory) and all other combinations). I have pasted the table with these data below so that you can see for yourself. The take home point here is that there were no significant differences between genders and the differences that might be there are different depending upon whether using perceived learning style preference or assessed learning style preference. I’ll spend more time on this later in this post.
 |
| From Dobson, 2010 |
For class performance, Dobson found no difference between undergraduate and graduate students (I would assume that grad students would get more difficult assignments and exams which makes comparing these groups difficult- if not, one would question the quality of graduate students that barely outscore undergraduates, no?). No gender differences were found in class performance.
For assessed sensory modality (by VARK), there were no differences found in class performance. With perceived sensory modality, the kinesthetic learners significantly underperformed compared to the rest of the class. Here again is another difference between the perceived learning style preference and what VARK found: as assessed by the VARK, no difference was found with kinesthetic learners.
The author lists the four main conclusions of the study as “1) Nearly two-third of the respondents correctly matched their perceived and dominant assessed sensory modality preferences, 2) there was a significant relationship between perceived sensory modality preferences and course scores, 3) there was no association between sensory modality preferences and status, and 4) there was a nearly significant trend in sensory modality preferences and sex.”
Three of the four conclusions are questionable. Conclusion #3 is ok in my book. Conclusion #1 is a bit misleading, as a large portion of those that “matched” their perceived and assessed learning style preferences had been classified as a VARK and thus would have matched with ANY perceived learning style preference. Conclusion #2 is supported by the data. In fact, the author discussed a previous study that he had conducted which found similar low scoring K learners in physiology classes. But why is this only found in those that perceive themselves as a K learner and not in those that are assessed as a K learner? This serves as further disagreement between VARK assessed learning style preferences and the perceived learning style preferences (more evidence against conclusion #1).
Dobson suggests that teachers may need to focus on K learners in physiology classes and review their teaching styles and practices to ensure that these learners are being helped. Another interpretation comes from other evidence which suggests that those that perceive that they have a learning style preference that is not in line with a teacher’s style of teaching may assume that they are going to do worse and simply not try as hard. Thus, they may be self-handicapping themselves (Reiner, 2010-2011). It is possible that the K learners may be doing this, and thus they will score lower. Thus, the classification of these students into a specific learning style preference may be detrimental to their success in class.
For conclusion #4, the data is quite messy. Statistically speaking, the difference between learning style preferences for males and females is not significant, but it is close. However, a comparison between gender differences between the perceived preferences and the assessed preferences shows no real pattern. Again, the disparity between the perceived preferences here and the assessed preferences provides more support against conclusion #1, but also makes interpretation of conclusion #4 impossible.
Overall, this study doesn’t really provide evidence for or against learning styles; that wasn’t their purpose. However, the data from this study question the validity of the VARK as an assessment of learning style preference.
References
Dobson, J. (2010). A comparison between learning style preferences and sex, status, and course performance. Advances in Physiological Education, 34: 197-204.
Riener, Cedar. Learning Styles: Separating Fact and Fiction. Psychology Teacher Network from the American Psychology Association Education Directorate. Winter 2010-2011. Vol 20, Issue 4.
No comments:
Post a Comment