A new meta-analysis of learning styles models from Turkey has some interesting findings about these types of educational interventions. Hare are key highlights from this research. To view the entire research study, you can use this link to view the full PDF File. This research article was published Online in October of 2016 in the journal Educational Sciences: Theory & Practice. 

https://pdfs.semanticscholar.org/f84c/50ad52ba6fee229ebad00fa4195af909ccc1.pdf

Overview of the meta-analysis

The total sample size included in this review was 2159. The treatment groups consisted of 1075 and the control groups consisted of 1084 students. Here was analysis of how the research studies were selected or inclusion in this meta-analysis.

The table below shows the frequency distribution of the studies by course type, study type, learning style model, experimental design and investigated variable.

According to Table 1, in terms of course type, 30% (f = 9) of the studies involved natural science (science, physics, chemistry, biology), 23.3% (f = 7) involved math, 20% (f = 6) involved social science (social studies, history, geography), 16.7% (f = 8) involved English, 6.7% (f = 2) involved informatics and 3.3% (f = 1) involved music. In terms of study type; 53.3% (f = 16) of the studies are thesis and 46.7% (f = 14) are articles.

In regards to the types of learning style models involved, here is the breakdown:

• 43.3% (f = 13) of the studies employed the Perceptual Learning Styles Model,
• 33.3% (f = 10) employed the 4MAT Model,
• 6.7% (f = 2) employed the Kolb Learning Styles Model,
• 10.1% (f = 3) employed the Dunn & Dunn Learning Styles Model,
• 3.3% (f = 1) employed the Felder & Solomon and Brain 2070 EDUCATIONAL SCIENCES: THEORY & PRACTICE Dominance Model.

In terms of experimental design, 86.7% (f = 26) of the studies used quasi-experimental design and 13.3% (f = 4) used true-experimental design. In terms of investigated variable, 58.0% (f = 29) of the studies investigated the effect of learning style on achievement, 22.0% (f = 11) investigated the effect of learning style on attitude and 20.0% (f = 10) investigated the effect of learning style on retention.

Effect Sizes of Learning Styles Models 

Under the random effect model, the common effect sizes were as follows.

Studies employing the…

• 4MAT System, 1.168 (0.860, 1.477);
• Perceptual Learning Style Model, 0.870 (0.653, 1.023);
• Dunn & Dunn Learning Style Model, 1.331 (1.047, 1.087);
• Kolb Learning Styles Model, 1.067 (−0.876, 3.009).

These results reveal that there is not a meaningful difference between the effects sizes of the classification created according to course type: namely the effect sizes of the courses tailored to learning styles model on the academic achievement are independent of course type.

When the effect sizes of 11 studies included in this review were combined according to the random effect model, the common effect size was calculated as (d) 1.113 with 0.227 standard error and 95% confidence intervals of 1.557 and 0.669. This common effect size is large according to the Cohen’s (1988) classification. This d value is associated with a U3 value of 84.1%, which means that the average student receiving instruction tailored to their learning styles scored higher on attitude tests than 84.1% of students receiving no instruction.

When the effect sizes of 10 studies included in this review were combined according to the random effect model, the common effect size was computed as (d) 1.763 with 0.241 standard error and 95% confidence intervals of 1.763 and 0.817. This common effect size is large according to the Cohen’s (1988) classification.

This d value is associated with a U3 value of 90.3%. This means that the average student receiving instruction tailored to their learning styles scored higher on retention tests than 90.3% of students receiving no instruction.

When the effect sizes of 10 studies included in this review were combined according to the random effect model, the common effect size was computed as (d) 1.763 with 0.241 standard error and 95% confidence intervals of 1.763 and 0.817. This common effect size is large according to the Cohen’s (1988) classification.

This d value is associated with a U3 value of 90.3%. This means that the average student receiving instruction tailored to their learning styles scored higher on retention tests than 90.3% of students receiving no instruction.     

In these meta-analytic review, it was found that when student received an instruction tailored to their learning style, their attitude (d = 1.113) toward courses improved and their learning was more permanent (d=1.290).

When comparing this to Hattie’s Barometer of Influence,  it is clear that this meta-analysis shows a very positive impact for learning style interventions and the 4MAT Model of Instruction in particular. The Chart above displays the impact of 4MAT on the Hattie Scale and places it near the top of the Desired Effects for an Educational Intervention. 

According to these results, it can be said that the learning environment designed based on learning styles has a large effect on students’ attitude toward course and learning retention.