We’ve all seen the diagram above, or a variant on it, in contexts ranging from articles for parents to tips for public speakers. Schemes of three, four, seven, eight, or any other number of “learning styles” are touted in popular culture as tools for career development, relationship enhancement, business success, and even spiritual growth. But do learning styles have any basis in fact, and can they really help students learn better?

In 2009, four cognitive psychologists were commissioned to evaluate the evidence to support using learning styles in instruction. Their conclusions, published in the article “Learning Styles: Concepts and Evidence,” were that (1) very few studies had used a research design capable of testing the validity of learning styles as applied to education, (2) of those studies that did, several found results that contradicted the effectiveness of using learning styles in education, and (3) “limited education resources would better be devoted to adopting other educational practices that have a strong evidence base.” As do all researchers, the authors also concluded that further research on the use of learning styles in education may be a good idea.

Accordingly, in 2015, a University of North Georgia assistant professor named Joshua Cuevas published “Is Learning Styles–Based Instruction Effective? A Comprehensive Analysis of Recent Research on Learning Styles,” a review of the research that had come out in response to the challenge issued in 2009. Before turning to what Cuevas uncovered, though, let’s look at some background on learning styles.

  • What does the phrase “learning styles” mean?

The idea of learning styles is the concept that different people have different “preferred,” or more natural, ways of processing information. Further, it assumes that if people receive instruction in their preferred way, they learn more effectively. This is referred to as the “matching hypothesis.” Various inventories have been developed to try to identify an individual’s learning style.

Where did the idea of “learning styles” come from?

Different researchers trace the origins of the learning styles idea to different sources, such as the Myers-Briggs assessment—which, incidentally, also lacks research/evidence to support its use (https://en.wikipedia.org/wiki/Myers%E2%80%93Briggs_Type_Indicator#Criticism). Other researchers point to multiple intelligence theory, which posits eight forms of intelligence—visual-spatial, verbal-linguistic, logical-mathematical, bodily-kinesthetic, interpersonal, intrapersonal, musical, and naturalistic—but also lacks supporting research or evidence (https://en.wikipedia.org/wiki/Theory_of_multiple_intelligences#Lack_of_empirical_evidence).

  • Learning styles inventories

An Internet search turns up thousands of quick questionnaires, or inventories, aimed at helping teachers characterize individual student learning styles, or older students label themselves. Some are free, others for sale. The inventories that appear most in the research are those developed by David Kolb (https://en.wikipedia.org/wiki/Kolb%27s_experiential_learning), which classify people into four categories: “divergers,” who favor feeling and watching; “assimilators,” who favor thinking and watching; “convergers,” who favor thinking and doing; and “accommodators,” who favor feeling and doing. The inventories most commonly used in practice, however, are based on a visual/auditory/kinesthetic (VAK) or visual/auditory/read-write/kinesthetic (VARK) classification scheme.

Several studies have identified problems with the validity and reliability of these inventories. Moreover, many researchers have pointed out the incongruity of a highly profitable and thriving industry based on learning styles, despite the lack of a systematic tool for their measurement or evidence for their use in instruction.

  • What do general teacher education books say about learning styles?

Most of the general teacher education textbooks that Cuevas reviewed in his 2015 comprehensive analysis advised teachers to take learning styles into account when planning their instruction, despite the fact that some of these texts were published after the 2009 review that found no evidence to support the theory of learning styles. Some of the texts conflated the (unsupported) theory of learning styles with the (also unsupported) theory of multiple intelligences, as if they are one and the same.

Many of these texts also tried to support their recommendation for incorporating learning styles in instruction by suggesting that teaching to learning styles is associated with increased academic achievement. Almost none of them, however, cited any peer-reviewed research to support this claim. Most simply stated it as if it were widely accepted.

The few studies on learning styles that any of these textbooks cited did not measure student learning outcomes. Although they used surveys to figure out students’ preferred learning styles, they stopped short of examining whether those students’ learning improved as a function of being taught in accordance with their identified learning style. In other words, none of the teacher education textbooks that recommended the use of learning styles in teaching offered any empirical evidence that the practice works.

In contrast to general teacher education textbooks, undergraduate educational psychology texts that Cuevas reviewed tended to describe the lack of evidence to support the use of learning styles in instruction.

  • The review of research published after 2009

For his 2015 review, Cuevas searched about 1,400 articles with “learning styles” in the title and identified 31 that examined the concept of learning styles as described above. However, more than half of these (16) came from “pay-to-publish” journals; Cuevas excluded these because of the questionable standards such “predatory journals” typically have. He noted, however, that almost all of them reported “positive” findings despite the fact that they had not been designed to answer the fundamental question of whether learning styles–based instruction could boost student learning—that is, most of them did not include an instructional intervention (a change in teaching method) based on students’ identified learning styles.

A few of the studies published in reputable journals were likewise based on correlations between learning styles and other variables, such as gender or the results of a different survey. None of these correlational studies set up an experiment to test the effectiveness of instruction based on identified learning styles.

Cuevas classified the remaining studies as experimental, though he cautioned that the quality of their methods and analyses varied. For one thing, most of them (indeed, most of the original 31 “learning styles” articles) used college students as the study population, even though the majority of instruction based on learning styles takes place in K–12. Let us take up, in turn, the experimental studies that supported the learning styles hypothesis and those that did not.

  • The studies that supported the learning styles hypothesis

A 2012 study looked at 70 first-year undergraduate medical students in Saudi Arabia, classifying them into either an “active” or a “reflective” style. Different instructional modes were incorporated into the teaching methods: independent study, group interaction, reasoning/problem solving, and active participation. The results showed clear differences in the behavior of students in the different categories (e.g., active learners communicated more during group work) but found no differences in overall learning. This study did not test the matching hypothesis (that if the type of instruction matches a student’s preferred style, the student will perform better academically).

A 2010 study’s authors created their own inventory to measure learning styles. They collected data on 64 undergraduates and divided them into two groups: one group learned via instruction intended to match students’ learning styles and the other via instruction that mismatched their learning styles, as measured by the authors’ inventory. The matched group showed increased learning efficiency (needing less time and fewer resources) but did not show better gains in achievement. The article did not include any data on the quality (reliability or validity) of the inventory, details of the analysis, or results of the academic assessments.

A 2012 study attempted to test the matching hypothesis with 98 Taiwanese college students, focusing on two styles (“input-oriented” and “perception-oriented”) and employing a control group with “unidentified” learning styles. The results were said to support the hypothesis, but Cuevas reported many concerns with the study: the paper omitted important information about how the study was conducted, very little was presented about the sample or teaching methods used, it was unclear whether the experimental groups contained students with a single learning style or students with various learning styles, and the term “unidentified” was not defined.

A 2011 study categorized 39 Taiwanese 5th grade students into two learning styles (“active” or “reflective”). The study took place during a single one-hour lesson, in which the instruction was identical for both groups except for the last 15 minutes, in which the active group brainstormed together and the reflective group received instruction and prompts before being asked to summarize. The study scored performance on an essay-type question and found that reflective learners’ gains were higher (i.e., they demonstrated more learning) when taught via reflective means (the instruction and prompts), and that active learners’ gains were higher when taught via active means (such as brainstorming). The authors stated, “matched groups learned significantly more than mismatched groups.” Cuevas suggested that, given the small sample of 39 students, this study should be replicated with a larger sample and over multiple lessons to see if the results are consistent.

Finally, Cuevas described  a 2012 study as “perhaps the strongest and most interesting research on learning styles,” though it did not look at academic learning. The study tested 151 elementary school students from four schools in the United Kingdom for sensitivity to sensory cues (https://en.wikipedia.org/wiki/Sensory_cue). It found that students’ learning styles as identified in the VAK system were related to their learning choices. Cuevas suggested these results may have value in assisting students with choosing academic courses or occupational tracks.

  • The studies that refuted the learning styles hypothesis

A 2009 study collected data on 70 third-year anesthesiology students from a dental school in South Korea. Researchers tested for an interaction between the students’ learning styles and achievement as measured by written responses to case problems. The study found no interaction effects, and the authors concluded it was more effective to have students adapt to different learning environments than to make sure instruction conforms to learning styles.

A 2010 study tested the matching hypothesis by looking at the performance of 33 college-level psychology students. It found no significant differences in gains in achievement based on learning style.

A 2011 study examined how learning styles impact course selection and achievement using data from 161 college freshmen, some taking a computer science class online and others taking it in a traditional format. The results suggested that learning style was unrelated to the course selection (online versus in-person) and that there were no significant differences in academic performance as a function of learning style and method of instruction.

A 2010 study tested the matching hypothesis using data from 159 undergraduate students using two different web-based instruction modules (one text-only and the other with multimedia and interactive components). The results suggested no difference in academic performance as a function of learning style and type of instruction.

A 2009 study followed 99 undergraduate students over three years and found that although students self-identified their preferred learning style consistently over time, groups that should have performed better on certain types of tasks based on their learning style did not perform any better on those tasks than students whose learning style did not match the type of task.

Finally, a 2011 study gathered data from 60 undergraduate and postgraduate students in Australia. It used a multimedia approach to deliver six experimental conditions reflecting slightly different instructional interventions. The authors concluded that there were no significant differences in learning across the groups based on their learning style and type of instruction. However, based on student feedback, they suggested that learning styles–based instruction would be beneficial by improving students’ motivation.

Cuevas pointed out that, curiously, some of the researchers whose findings refuted the learning styles hypothesis continued to argue for using an instructional approach informed by learning styles even though their data appeared to contradict this conclusion.


Overall, Cuevas’s 2015 review concluded that although research is always warranted, the existing research suggests that it is irresponsible for teacher education programs and public educators to apply the theory of learning styles in practice, especially when it often means taking the focus off of interventions and strategies that have stronger evidence and are more likely to help learners.