Chapter Two

An Investigation of Learning Styles in General Chemistry Students


A number of models have been proposed to describe differences in how individuals take in and process information. The concept has been approached by both pedagogists and cognitive psychologists, with apparently little coordination of effort or intentional sharing of information. Both types of researchers, however, agree that such differences exist. In this review of the literature and historical bases, the major models are described and compared with the Jungian approach used in this research.

The review of the literature was partially accomplished using the United States Office of Educational Research and Improvement (OERI) ERIC system and ERIC Document Reproduction Service, with searches of various combinations of Jung, learning, styles, science, chemistry, secondary, cognitive, cognition, Dunn, Kolb, Bloom, Perry, Hansen, Silver, Strong, Myers, and Briggs, and other writers encountered in the search. Additional materials were located by physical searches of the library shelves, including current and back issues of education and psychology journals, reference materials and texts, texts used in the course work for the degree; physical searches of the education and psychology sections of the Barnes and Noble bookstore in Greenville, South Carolina and many other bookstores from 1992 through 1996; personal purchase of a number of Jung's works in translation and others' writings on Jung's work and that of other learning theorists; membership in psychology book clubs and the American Psychological Society, and Internet searches and discussion groups for additional leads. I have purchased most of the books and materials listed in the bibliography so as to have them constantly available for extended reading and study.


Dunn and Dunn have defined detailed descriptions of "learning styles" that describe environmental conditions and sensory preferences, and have labeled learners as either analytical or global. Their work has been so extensively published and widely disseminated that the term "learning styles" has become synonymous with their descriptors of the modalities of auditory, visual, and tactical-kinesthetic; environmental preferences for lighting, sound, food and comfort; and the analytical or global labels. Dr. Rita Dunn adheres strictly to her own model, and eliminates as unreliable or invalid all but three "comprehensive models." In her Survey of Research on Learning Styles (1994, part 2)Dunn dismisses any models that do not match her own, thus warping the literature to her own purposes. She specifically writes "others address only one to four elements, usually on a bipolar continuum," (1994, part 2) and dismisses those models. Dunn and Dunn direct the Center for the Study of Learning and Teaching Styles at Styles at St. John's University and travel and lecture extensively at in-service workshops for school districts across the country.

Problems with their research became obvious to this author, upon personal conversation with Dr. Rita Dunn, at a local "in service" for teachers (personal conversation, 10/21/94). Dr. Dunn supported her theses on learning styles by describing numerous studies conducted by her graduate students, in their classrooms, with small samples of students and questionable research methods (non random sampling, uncontrolled for ability or other similar factors, short term studies, uncontrolled for teacher differences, etc.) She described placing students in either an analytical environment or a global environment, with pre- and post testing to determine efficacy of environmental changes. The students were previously designated as analytical or global by the behavioral descriptors or preferences prescribed by Drs. Dunn. Efficacy of environmental changes were measured by comparison of scores of the students in each environment, analytical or global.

The environment preferred by the analytical student is reported to consist of straight and unpadded furniture, structured as standard desks and chairs, bright lighting, quiet and free of disturbances. In all, the analytical environment is what has long been considered the ideal study environment. The environment preferred by the global learner is reported to consist of comfortable seating, in the form of couches and armchairs, cushions and carpeted floors. Lighting level is reduced for a softer effect, and music, food and drink are permitted. Learners are encouraged to find their own space or niche, wherever they prefer in the room, and relax comfortably while they work on their studies. Dr. Rita Dunn then reported that the students in the global environment are told if they don't do better in the global environment than they did in the regular (analytical environment) classroom, they would have to go back into the other room. Such priming of students to work harder in order to keep the perceived privileges of comfortable surroundings invalidates the research that serves as the basis for Dunn and Dunn's theories. What is established by such conditioning is only that students will work harder for perceived privileges, or bribery works.

A study, which is typical of the literature on Dunn's model (Dunn, 1990, included in Dunn, 1994, part 2), intended to describe differences in preferences by ethnicity, uses groups of students in widely diverse geographic areas. The study populations were African Americans from suburban Suffolk County New York, Chinese Americans from Chinatown in lower Manhattan, New York City, Greek Americans from an urban parochial school in Queens, New York, and Mexican Americans from a rural school in La Joya, Texas. These widely diverse groups, with nothing else in common except that they were all either elementary or middle school students, were compared for learning environment preferences, and group differences were attributed to ethnicity alone. The research did not consider that these students had different teachers, content, texts, school buildings and districts, and everything about their education and living environment was different from the other groups. It is clearly invalid to consider pre and post test differences to be due solely to the controlled variables of learning preferences. The body of research on Dunn's work is conveniently collated into the two volumes cited and sold by the Dunns' Center at St. John's.

There is no doubt that some learners prefer a room warmer or cooler, brighter or less bright, prefer quiet while studying or prefer to study with music playing. That these environmental preferences influence grades has not been established in a large, appropriately controlled, statistically rigorous study.

Beyond the obvious problems with the research, I believe Dunn and Dunn's theories have some basis in reality, and it is easy to superimpose Dunn's model upon the Jungian. The descriptions Dunn uses of the analytical learner coincide well with the Jungian model ST learner. That leaves the global descriptor for all other learners. Dunn (1994, teacher's in service workshop) reported that about twelve to eighteen percent of the population is analytical, and up to 88 percent are global in learning style. Rita Dunn reported herself to be an analytical learner. She claimed there is no judgment attached to the designations analytical or global, but she described analytical learners as organized, responsible, and dependable, and global learners as disorganized, irresponsible and undependable.

Personality Types

Katherine Myers and Isabel Myers-Briggs (Myers, 1958, 1962) (see Howard, 1992, for an example of use of the Myers-Briggs MBTI) have developed Jung's work in psychological type into an instrument of 126 questions, of which in this author's opinion (and that of many test takers of the author's personal acquaintance) many are ambiguous. "Which mistake would be more natural for you: (a) to drift from one thing to another all your life, or (b) to stay in a rut that didn't suit you?" (Briggs, 1991)

The scoring of the completed instrument is accomplished by the Myers and Briggs organization at a cost to the test taker, and results in the individual's being classified as one of sixteen types. The subject is scored on Jung's four bipolar descriptors: introversion vs. extroversion, sensing vs. intuitive, thinking vs. feeling, and judging vs. perceiving. The model is based closely on Jung's psychological types, but expands the application of theory to practice with the instrument. Scoring the responses to the 126 questions determines whether an individual is introverted or extroverted, sensing or intuitive, thinking or feeling, or judging or perceiving. No degree of either extreme is expressed nor implied. If an individual chooses answers that very slightly more often expresses extroversion than introversion, the result reported is that the individual is extrovert. Likewise, if the individual consistently chooses extroversion responses over introversion responses, the result reported is extrovert. There is no expression of degree, although the two subjects may be very different in this characteristic. Dunn and Dunn's figures of 12-18% analytical and 82 to 88% global coincide poorly with results of years of cumulative Myers-Briggs research, which is reported by Tieger (1992). Tieger examines career choice by personality type, under the Myers-Briggs model, and collates years of research with the Myers-Briggs MBTI.

Extensive investigation undertaken for the New York Board of Regents looked at the role of culture, left vs. right brain theory (see also Williams, 1983) and environmental factors in addition to cognitive differences. The major conclusions of that board were that students do have major learning styles differences, and that designing and using multiple instructional strategies to teach to these styles can improve teaching and learning. (Yale, 1988)

Perception and Process

Kolb (Yale, 1988) saw the model of differences as two bipolar descriptors, crossed as the Jungian model used in this research, the Kolb descriptor for perception being concrete to abstract, oriented in his model vertically with the concrete at the top and abstract at the bottom, (similar to Jung's model) intersected laterally by the process bipolar, from doing, at the left, to watching, on the right. (In contrast to Jung's model of thinking to feeling.) Kolb also emphasized all types are valuable and complement one another. Basing his premise on the work of Lewin, Piaget and Dewey, Kolb also used a four quadrant model, in which he described imaginative, analytical, common sense, and dynamic learners. The Regent's report examined a wide variety of models, many of which are very similar. Education researchers are closing in on good models to describe and categorize learners. It remains for the educators to assimilate and apply the appropriate variety of teaching strategies in the classroom, and for the students to shape their study efforts to best fit their strengths in perception and processing.

Hemispheric Dominance

Many researchers have studied brain hemispheric dominance (Williams, 1983). Although arguably one of the more controversial issues, as far as actual physical connection between certain information processing preferences and strengths and left or right brain hemispheres, the groupings of patterns of behavior and cognitive activities is helpful to the educator. The left brain is described as "serial, analytic, rational and verbal," while the right brain is "global, visual and holistic" (Yale, 1988, after work published by neurologist Dr. J. Bogen in 1975). To connect hemisphericity with Jung it would seem that the left hemisphere mode is associated with the sensor, and the right hemisphere mode is associated with the intuitive, with some overlap and some discrepancies.

Agreements and Similarities Across Models

Most of the models reported in the literature can be fitted to one another, which is only logical since all are attempting to describe the same widely diverse population. It is the very diversity of the student population that raises controversy. Certainly the age group and environment of residence of the student study population causes wide shifts in data and conclusions. Most researchers agree (if such a broad statement can ever be made) that there are differences in how people learn, and a variety by some descriptor of teaching strategies would be preferable and more effective. A difficulty lies with some educators and administrators, who are perhaps Jungian described dominant ST learners, seeing issues as black or white, who tend to believe that there is only one way to do things, their way, and who typify the population of teachers. ST's historically do well in school by memorizing content, learning through repetitive drill, and using concrete manipulatives to learn and to teach. Since these practices work so well for themselves, ST educators tend to believe strongly that it must therefore work best for everyone. A conclusive mathematical approach to identification and description of differences in how individuals acquire and process information is necessary to convince this population of ST dominant educators. This research is intended to fill that need.

Much research has been conducted into finding the one best method of teaching for any one course, with varying results. Any new technique can be shown to improve scores for some students in a group. This phenomenon should demonstrate that no one method is best for all learners. Whether it is new math vs. "old math," cooperative learning vs. individual effort, cookbook labs vs. discovery learning, whole language vs. phonics, or manipulatives, or constructivism, each method will work better for some students than for others. Because of the nature of student populations, class sizes, and the typical geographical considerations of school populations, variation in results is statistically supported. Educational studies have been dominated by small populations, poorly controlled variables, and less than rigorous statistical work, and recently educators have tended to perform qualitative studies and to pursue a subjective "I know it works because I know my students," argument. As educators we know what works for us, however, it is still necessary to apply rigorous statistical analysis, large populations, and carefully controlled experimental techniques to develop new knowledge that can be transferred to populations other than the sample studied. This study seeks to fulfill that requirement.