Share this article on:

Learning and Cognitive Styles in Web-Based Learning: Theory, Evidence, and Application

Cook, David A. MD


Cognitive and learning styles (CLS) have long been investigated as a basis to adapt instruction and enhance learning. Web-based learning (WBL) can reach large, heterogenous audiences, and adaptation to CLS may increase its effectiveness. Adaptation is only useful if some learners (with a defined trait) do better with one method and other learners (with a complementary trait) do better with another method (aptitude–treatment interaction).

A comprehensive search of health professions education literature found 12 articles on CLS in computer-assisted learning and WBL. Because so few reports were found, research from nonmedical education was also included. Among all the reports, four CLS predominated. Each CLS construct was used to predict relationships between CLS and WBL. Evidence was then reviewed to support or refute these predictions.

The wholist–analytic construct shows consistent aptitude–treatment interactions consonant with predictions (wholists need structure, a broad-before-deep approach, and social interaction, while analytics need less structure and a deep-before-broad approach). Limited evidence for the active–reflective construct suggests aptitude–treatment interaction, with active learners doing better with interactive learning and reflective learners doing better with methods to promote reflection. As predicted, no consistent interaction between the concrete–abstract construct and computer format was found, but one study suggests that there is interaction with instructional method. Contrary to predictions, no interaction was found for the verbal–imager construct.

Teachers developing WBL activities should consider assessing and adapting to accommodate learners defined by the wholist–analytic and active–reflective constructs. Other adaptations should be considered experimental. Further WBL research could clarify the feasibility and effectiveness of assessing and adapting to CLS.

Dr. Cook is assistant professor of medicine, Mayo Clinic College of Medicine, Rochester, Minnesota.

Correspondence should be addressed to Dr. Cook, Baldwin 4-A, Division of General Internal Medicine, Mayo Clinic College of Medicine, 200 First Street SW, Rochester, MN 55905, e-mail: 〈〉.

Note: References 76–89 are cited in Table 1 and Table 2 only.

Cognitive and learning styles (CLS) have been a topic of research and discussion in medical education for decades. Uses—either suggested or demonstrated—of these learner characteristics have dealt with nearly every aspect of the education process.1 With the development and implementation of innovative educational technologies, assessment and use of CLS may find new and even more important roles. Web-based (WB) learning makes adaptation to the individual learner more feasible than in the past, and research in this field may be more productive than prior research in CLS. In fact, adapting WB environments to specific CLS—replicating the actions of an effective teacher who adapts instruction to meet the needs of individual learners in face-to-face (FTF) teaching—may be necessary to maximize the benefit from WB learning.2 Such adaptation is particularly important given the large and often heterogenous body of learners a WB course may reach.3

In considering adaptations based on CLS, the concept of aptitude–treatment interaction4 is critical. An interaction occurs when a learner attribute predicts a different outcome depending on the teaching method. For example, interaction is present if group 1 (with attribute 1) learns better with method A than method B, while group 2 learns better with method B than method A. Adapting the method so that group 1 gets method A and group 2 receives method B would enhance learning. If the teaching method improves performance without interaction (e.g., both groups do better with method B), this implies that all learners (regardless of attribute) will benefit, and adaptation is thus unnecessary.

In this review, I discuss the theory and evidence underlying potential applications of CLS to WB learning, with a focus on aptitude–treatment interactions. For each CLS, I will predict relationships to WB learning based on the theory or definition of the construct or dimension (I have used these terms interchangeably), examine the literature for evidence to support or refute these predictions, and summarize the theory and evidence to propose applications of CLS to WB learning.

Back to Top | Article Outline


Web-based learning

WB learning is a subset of computer-assisted learning (CAL), in which instruction is accomplished primarily or solely using the Internet or a local intranet using hypertext-based information.5,6 Although it has much in common with other CAL modalities, WB learning is unique in at least three respects. First, it allows near-instantaneous communication and access to resources between distant sites. Second, the information constituting the course itself resides in a central location, which facilitates content updates and course modifications. Third, the programs are relatively technology-independent, and will usually function and have a similar appearance regardless of the local computer configuration. Other features commonly found in WB learning, such as hypertext links and multimedia, are also found in other forms of CAL. Because of these common features, many studies of non-WB CAL are relevant to the present discussion.

Although “WB learning” is often discussed as a unitary construct,7–9 this is not the case. The common thread in WB learning is the use of hypertext and the Internet, but this leaves room for wide variation in configuration, instructional method, and presentation. Indeed, these details often form the basis of proposed adaptations. Many reports fail to adequately describe these distinctions (as Tables 1 and 2 illustrate), even though this information is essential for interpretation and application of research.

Table 1

Table 1

Table 1

Table 1

Table 2

Table 2

Table 2

Table 2

Table 2

Table 2

WB learning environments can be divided broadly into discussion-based and Web-page-based “configurations.” Discussion-based learning parallels the configuration of a small-group learning session. Learners communicate with each other while the instructor provides guidance, learning activities, and resources as appropriate. Compared to FTF groups, WB interactions are different:10,11 discussion is often asynchronous (with lapses in time between communications), and although it is a group process, the learner works alone. Page-based learning parallels the configuration of a lecture, as the learner “attends” a Web page that “delivers” content and activities prepared by the instructor. Neither configuration is inherently better or worse than the other. Rather, just as small-group learning and lectures each have a role in FTF teaching, discussion-based and page-based configurations have strengths and weaknesses that make each appropriate in different circumstances. Elements of both configurations are often incorporated into a single course.

Instructional method is a critical but often overlooked aspect of WB learning.12,13 Just as in FTF teaching, WB instructional methods (e.g., case-based learning, self-assessment questions, group projects, simulation) play an important role in the success of the learning activity.14 Although discussion-based and page-based configurations are methods of instruction, their distinction as a separate level of variation is supported both conceptually (configuration is frequently defined before instructional methods are selected in both WB and FTF instruction) and logistically (discussion-based learning requires different planning than page-based WB learning).14

The third level of variation is presentation. Nearly all WB environments (whether page-based or discussion-based) are primarily text, but most employ “enhancements” such as multimedia (e.g., pictures, sound, or video), interaction between learner and page (hyperlinks to other resources, interactive models and games), or high simulation fidelity. A site consisting only of text will likely have a different effect on learning than a site enhanced by activities and multimedia, and this may be affected by (interact with) CLS. Web site structure and navigation are other important factors in presentation.15

Back to Top | Article Outline

Cognitive and learning styles

Curry16 proposed a tripartite model of cognitive and learning style in 1983 that continues to provide structure to these diverse constructs. The first (outer) layer is instructional format preference, an “affinity for various modes of information delivery or access”17 such as environmental (presence of sound or light) and social (alone or with peers) conditions. These preferences are easily altered and thus relatively unstable. The middle layer is learning style—“conscious and intentional strategies that individuals employ to achieve well-defined ends... most often observed as individually consistent orientations to learning and studying.”17 These attributes are more stable and resistant to outside influence. The inner core is cognitive style—a student’s pattern of “perception, memory, thinking, and judgment.”17 These characteristics are stable and may in fact be lifelong attributes, relevant to both learning and nonlearning contexts. In this article, I will be concerned only with the latter (inner) two types of style.

Multiple labels have been proposed for constructs of cognitive and learning styles, and many are redundant.1,18,19 I have attempted a reconciliation, outlined in Table 3, that groups constructs with similar definitions under a single label. Once a construct has been defined in this article one label will be consistently used to refer to all constructs represented by that definition regardless of the specific model or instrument.

Table 3

Table 3

Back to Top | Article Outline

Literature Review

Medline and ERIC were searched using the terms learning style, cognitive style, or learning preference, and computer-assisted instruction, computer-aided instruction, computer-assisted learning, computer-aided learning, Internet, Web, or online. In ERIC the results were limited to post-secondary education, two-year colleges, and graduate education. Abstracts were reviewed, and articles presenting original data regarding CAL and CLS were selected for further review. Bibliographies were reviewed for additional citations. Unpublished abstracts and dissertations were excluded. The 16 articles involving learners in health professions education are summarized in Table 1. Of these 16, four articles assessing only learning preferences or qualitative observations were excluded from further discussion, leaving 12 studies of CLS. Because of the small number of studies identified among the health professions, the search was extended to include other fields of adult education (Table 2). However, the review was not exhaustive beyond health professions education.

Four constructs of CLS dominated the studies reviewed, and are described below. For each style, relationships to WB learning were predicted based on the construct definition. Evidence was then sought to support or refute these predictions. Marked heterogeneity of learners, settings, interventions, and research methods precluded pooling of results for meta-analysis.

Back to Top | Article Outline


The wholist–analytic dimension

In 1991, Riding and Cheema18 reviewed the literature on cognitive styles and concluded that most of the numerous labels were redundant. They proposed that all cognitive styles could be unified into two dimensions: the wholist (i.e., “whole-ist”)–analytic dimension and the verbal–imager dimension. Rayner19 repeated this work and arrived at similar conclusions.

The wholist–analytic dimension characterizes the manner in which an individual perceives the world–as a unified whole, or as a collection of individual parts. Because these are patterns of “perception [and] thinking”17 each constitutes a cognitive style. When perceiving information, individuals who are wholists will see “a balanced view of the whole,” while analytics “will separate it out into its parts.”20 In social situations, the wholist sees the group as a whole, while the analytic sees the group as a collection of individuals. Wholists benefit from structure, guidance, feedback, extrinsic motivation, and social activity in their learning,2,21 while analytics need less of each of these.

This construct subsumes most of the previously defined cognitive styles18,19,22 including field dependence/independence, impulsive/reflective, leveler/sharpener, diverger/converger, and holist/serialist. Some authors have argued that field independence is an ability (i.e., more field independence is better) rather than a style,19 but this argument has not been specifically extended to the wholist–analytic construct. Nonetheless, wholist–analytic appears to be the only construct for which suggestions to improve learning consistently involve mismatching (e.g., providing structure for wholists) rather than matching (e.g., using images for visual learners).

As part of their Index of Learning Styles,23 Felder and Silverman24 proposed the sequential–global dimension of learning style (defined in terms of orientation to learning) that shares many features with the wholist-analytic dimension.* Sequential learners follow a linear (analytic) process, while global learners “learn in fits and starts... until suddenly they ‘get it.’ ”24

WB learning environments often lack a clear structure—especially those using extensive hyperlinks—and require internal motivation. In addition, page-based learning can be socially isolating. In these contexts, analytics (who can provide their own structure and require less external motivation and social support) should have an advantage over wholists. In settings with explicit guidance or structure, external motivation, and social interaction, wholists have the advantage.21 The social nature of discussion-based WB learning should initially favor the wholist, but as discussion threads become long and complex the analytic (who can extract salient themes from the milieu) would be favored.

The only study25 in health professions education to assess the styles defined by the wholist–analytic dimension in connection with WB learning or CAL found no influence of style on achievement or attitude toward CAL tutorials.

There are, however, many studies in other fields. Research evaluating instructional methods has found interaction between wholist–analytic styles and depth of studying, use of overviews, and structuring of information. Analytics studied material in greater depth than wholists,26,27 while wholists used an overview more often than analytics did.27 Wholists did better in a page-based WB environment that provided a broad overview before pursuing topics in detail, while analytics did better when exploring in detail first.28 Another study showed a similar trend,29 but a third study30 failed to confirm this. Wholists showed a trend toward improved performance when using a tool that made information structure explicit, while analytics did better when information was less structured31; however, another study found no interaction between style and a tool to facilitate structuring of knowledge.32 An adaptive WB system33 targeting all of these strategies at once was more effective than the control WB intervention. Contrary to my predictions, a structured, independent-learning CD-ROM favored analytics and an unstructured WB format with group discussion board favored wholists,34 but differences in motivation and social interaction between the two formats could account for these observations.

Looking at presentation, studies have found differences in navigation patterns, with wholists following a more direct path through the CAL or WB environment,26,27 achieving higher test scores using a broad-before-deep site structure,28 and doing better with short Web pages.30 Furthermore, analytics had higher test scores than wholists when using a WB environment with many hyperlinks, while there was no difference between styles when using a less complex CD-ROM.35 In contrast, one study36 failed to find significant interaction between style and hypertext structure, and wholists followed more hyperlinks in another study.37

Overall, the evidence appears to support the theory-based predictions. Analytics perform better in CAL or WB learning environments that are less structured or encourage studying in depth before presenting an overview. Wholists do better in environments that provide structure and a global perspective, and may benefit more from social interaction.

Back to Top | Article Outline

The verbal–imager dimension

Riding and Cheema18 also proposed a verbal–imager (verbal–visual) dimension of cognitive style. The verbalizer learns best from words (speech and written text), while the imager learns best from pictures, demonstrations, and displays. Imagers are also better at keeping track of their location in space. Riding20 suggests that this construct has social implications, with verbalizers better at communication and imagers more focused on a “world internal to the individual.” Imagers are also more likely to continue with tasks they find boring while verbalizers require stimulating presentations.38

Felder and Silverman24,39 proposed a similar visual–verbal dimension of learning style. Kirby et al.40 described a visual–verbal construct as well.

Verbal learners should learn better than imagers in most WB learning settings since text is the dominant form of information in both page-based and discussion-based configurations. However, imagers should do better in multimedia-rich environments and in complex WB environments where they will remember their “location” in cyberspace.

Evidence does not support these predictions. The sole study41 in health professions education assessing verbal–imager styles failed to find a persistent difference in performance among participants using a visually intensive CAL environment. This study was limited by a skewed population (nearly all the learners were imagers). Some studies in nonhealth professions have supported the predictions made above by showing that imagers prefer images42 and better remember their location in a complex WB learning environment,30 verbalizers do better on essays,30 and wholist verbalizers do better overall in CAL.29 However, several studies have failed to demonstrate significant findings,11,43,44 and others report counterintuitive results. For example, one study35 found that verbalizers performed better using a multimedia-rich WB environment than using a CD-ROM that had less multimedia, while imagers did better using the less-visual CD-ROM. Another study45 found that imagers posted more messages to a WB discussion board than verbalizers did, and that fewer verbalizers completed the discussion-based course.

The evidence is mixed—showing supportive, neutral, and counterintuitive results. Although this issue merits further study, it is possible that the verbal–imager dimension does not play a significant role in WB learning and CAL.

Back to Top | Article Outline

The concrete–abstract dimension

Kolb proposed an experiential learning theory in which “knowledge is created through the transformation of experience.”46 In this model, information must first be “grasped,” or acquired, and then transformed, or processed. Learners continually choose how they will acquire new information (concretely, through the senses, or abstractly, by analyzing) and transform it (reflectively, by watching and thinking, or actively, by doing). The preferred choice along each of these two dimensions (acquisition and transformation) constitutes a style (for example, a learner may prefer abstract acquisition and reflective processing). Because these styles are defined in terms of “individually consistent orientations to learning,”17 they are learning styles. Kolb47 developed the Learning Style Inventory to assess these styles. Although Kolb’s model and inventory have been criticized,48–50 others have defended them.51 The model and inventory continue to be generally accepted, and the Learning Style Inventory is one of the most widely used measures of CLS.52

The first dimension—information acquisition—has the poles of abstract conceptualization and concrete experience. The concrete learner learns best from specific examples or experiences, and relates well to other people.47 The abstract learner prefers a conceptual, analytical, systematic approach to learning, and works better with theories and ideas. Rezler53 defined a parallel dimension in the Learning Preference Inventory.

Jung54 defined a similar construct in terms of psychological types (patterns of “perception,... thinking, and judgment,”17 and thus cognitive styles according to Curry): the sensing–intuitive dimension, assessed by the Myers-Briggs Type Indicator.55 Sensors prefer “what is real” (facts, data, and experimentation), while intuitors look for patterns and meaning (principles and theories). Felder and Silverman24 proposed a parallel dimension of learning style. In their definition, they note that “Intuitors are more comfortable with symbols. Since words are symbols, translating them into what they represent comes naturally to intuitors and is a struggle for sensors.”24 Thus, there may be some overlap of this definition with the verbal-imager construct above.

Differences in instructional method should show interaction with concrete–abstract styles. For example, case-based learning or interactive demonstrations would favor concrete learners, while using data to introduce a principle or basic mechanisms to explain a disease would favor abstract learners. On the other hand, it seems unlikely that WB configuration or presentation will interact with concrete–abstract styles. Concrete learners may do better with the interpersonal interactions of discussion-based learning, but this would be offset by the extensive use of words (abstract symbols). One possible exception is that concrete learners may do better in a multimedia-rich environment, in which the media represent vicarious experiences. When comparing WB to noncomputer interventions, the impact of the concrete–abstract dimension is hard to predict, as effects would likely depend more on variation in instructional method than on properties of the media formats.

Of nine reports in health professions education examining the concrete–abstract dimension,56–64 none compared different CAL or WB instructional methods. One study56 found that abstract learners using interactive text and graphics had higher test scores than did concrete learners, but another with similar instructional methods found no difference.62 Test scores did not differ by style for a WB text/image tutorial64 or for a series of CAL simulations.59 Course grades were higher (no statistical test reported) for concrete–reflective learners (“divergers”) in a distance-learning WB course than for divergers in the FTF classroom setting.60

Looking outside of health professions education, instructional methods matching abstract or concrete style demonstrated improved outcomes compared to methods that mismatched style.65 On the other hand, studies of hypertext-rich CAL and WB learning37,66,67 found no significant difference in outcomes between concrete and abstract learners, although there was no comparison with another CAL format. A comparison of two navigation formats68 found no difference between concrete and abstract learners’ outcomes. One author69 reported that learning styles were more concrete following CAL compared to preintervention, but the CAL interventions are not described and the findings have not been duplicated.

In summary, these findings agree with the prediction that concrete–abstract learning styles would not interact with CAL or WB learning configurations or presentations. However, one report65 suggests that matching styles in this dimension with instructional method may be effective.

Back to Top | Article Outline

The active–reflective dimension

The second dimension of Kolb’s model offers the choice between active experimentation and reflective observation. Active learners prefer practical application of information they have received, and are willing to take risks as they influence situations and people.47 Reflective learners, on the other hand, internalize information—observing before passing judgment, examining from different perspectives, and looking for meaning as they create new knowledge. Felder and Silverman24 defined their active-reflective construct using Kolb’s construct.

Both Felder and Silverman24 and Kolb52 note substantial similarity between this dimension and the extravert and introvert personality types defined by Jung54 and assessed by the Myers-Briggs Type Indicator. The extravert focuses attention and energy outward to activity and people, while the introvert focuses attention inward to his or her ideas and experiences.55

Most WB learning would seem to favor reflective learners. The asynchronous and independent nature of WB environments allows the learner to proceed without haste, promoting reflection but possibly boring the active learner. Activities in page-based learning, such as moving from page to page or pursuing hyperlinks, may not fully compensate for this. Discussion-based learning is inherently highly reflective. An exception to the pattern favoring reflective learners would be instructional methods using highly interactive games and models, which may meet the needs of the active learner but fail to allow adequate time for reflection. It is thus expected that reflective learners will prefer and do better with WB learning compared to active learners except when interactive features are prominent.

Of eight studies in health professions education examining the active-reflective dimension,56–61,64,70 only one compared different CAL or WB interventions. This study70 found an interaction between instructional method and style, confirming the author’s hypothesis that reflective learners would learn better with linear CAL while active learners would learn better using an interactive program. In other research, active learners did better than those with any other learning style when using interactive CAL,56 but test scores did not differ by style for a WB text/image tutorial64 or for a series of poorly defined CAL simulations.59

Results from studies outside of health professions education are similar. One report65 comparing instructional methods found a trend suggesting interaction consistent with the above predictions. Another study34 hypothesized that active learners would do better than reflective learners when using a highly interactive WB format, while reflective learners would outperform active learners when using a minimally interactive CD-ROM. The results partially confirmed this hypothesis, as active learners did better with the WB format, but performance using the CD-ROM did not vary whether the learner had an active or a reflective style. Reflective learners had higher test scores than active learners using a “complex, abstract” WB intervention.67 However, another study found that active learners used hyperlinks more than reflective learners, and obtained higher test scores when they did.71 Furthermore, two studies suggest that learners assume a more active learning style with CAL69 or WB learning.72 A study comparing linear and nonlinear navigation formats68 and two studies37,66 using hypertext-rich CAL and WB learning found no difference in performance between active and reflective learners.

Thus, the prediction that reflective learners would do better than active learners in most WB environments is not supported by the literature. This likely results from a paucity of evidence and varying degrees of active and reflective learning among existing studies. Alternatively, it could be due to faulty reasoning in the prediction, or a flaw in the active-reflective construct itself. Further research could clarify this issue. Limited evidence does suggest an interaction between instructional method and learning style, in that active learners do better than reflective learners with interactive WB learning and CAL, while reflective learners do better with methods that promote reflection.

Back to Top | Article Outline


In summary, the theory and evidence presented above suggest that matching learners to specific WB formats, based on whether they have a wholist or an analytic style, has strong support in both theory and the literature as a means to improve performance. Specifically, analytics tend to do better with less structure, longer Web pages, and a “deep before broad” approach. Conversely, wholists do better with more structure, shorter Web pages, a “broad before deep” approach, and social interaction. This aptitude–treatment interaction suggests that WB learning could be significantly enhanced by adapting presentation and instructional methods to styles in the wholist–analytic dimension. The wholist–analytic dimension has received little attention in medical education research.

Evidence neither supports nor clearly contradicts the predictions made from the theory of the verbal–imager construct. Whether this is due to inadequate research or a faulty construct, or whether these styles are simply not relevant to WB learning, remains to be seen. At this point, adaptations to specifically target the verbal–imager dimension are not justified.

Regarding the concrete–abstract dimension, both theory and limited evidence agree that adaptations targeting specific instructional methods will have more impact on learning than changes to the configuration or presentation of a WB learning tool. Such adaptations have not received much study in WB learning or CAL, but have been the subject of research in FTF teaching. Adaptations based on FTF evidence would likely be effective, but further research should confirm their merit.

For the active–reflective dimension, theory and limited evidence suggest that active learners do better than reflective learners with interactive WB learning and CAL, while reflective learners do better with methods that promote reflection. This does not imply that reflective learners do not require learning activities, but rather that reflective and active learners may benefit from different activities.

This analysis is limited by the paucity of reports, small sample sizes, and heterogeneity of study populations, interventions, CLS models, and outcomes. Many studies suffered from methodological deficiencies, incompletely defined interventions, and (especially those in health professions) the absence of a clear hypothesis. It is also important to point out that although this article considered CLS constructs with similar definitions as a single dimension (see Table 3), this assumption should be viewed with caution until empirically verified.

This study illustrates that it is not sufficient to simply define an intervention as “Web based” when conducting and reporting research. Instructional methods, configurations, and details of presentation vary widely and are critically important. It is also notable that nearly all of the significant findings involved specific modifications of the CAL environment rather than comparison of CAL to a noncomputer format. Among health professions education, both studies41,70 comparing distinct CAL interventions found significant differences among CLS, while only one60 of the four studies comparing CAL to another media format found a difference. Most of the controlled studies in nonhealth professions education compared two CAL formats, and nearly all of these found significant differences among CLS. The pitfalls of comparing CAL to noncomputer media have been described73–75 and appear to hold true for CLS research.

Future research could use CLS theory and existing evidence to develop and test hypotheses regarding the role of CLS in WB instructional design. Experimental studies comparing interventions should seek to confirm predicted aptitude–treatment interactions, and might consider matching and mismatching learners based on CLS. Changes in interventions should focus on variations of WB design—configuration, presentation, and instructional method—rather than comparisons with noncomputer or no intervention. In addition, studies should further evaluate the theoretical basis of CLS, attempt to consolidate the numerous CLS models, and verify the validity of CLS assessments.

Back to Top | Article Outline


Aptitude–treatment interactions form the basis of evidence-based adaptations to any teaching method. In the case of CLS in WB learning, interactions have been demonstrated for styles in the wholist–analytic dimension and, to a lesser extent, the active–reflective dimension. Limited evidence suggests interaction between concrete–abstract styles and instructional method, but confirmation is needed. Research in CLS is limited by deficiencies in study design and a multiplicity of models.17 Yet the findings described above offer hope that, at least for WB learning, investigation and application may be productive. Further WB learning research could clarify the feasibility and effectiveness of assessing and adapting to CLS.

Back to Top | Article Outline


1 Curry L. Cognitive and learning styles in medical education. Acad Med. 1999;74:409–13.
2 Triantafillou E, Pomportsis A, Georgiadou E. AES-CS: adaptive educational system based on cognitive styles. Paper presented at the Adaptive Hypermedia 2002 Workshop on Adaptive Systems for Web-based Education, Malaga, Spain, 2002. Universidad de Malaga.
3 Brusilovsky P. Adaptive educational systems on the World-Wide-Web: a review of available technologies. Paper presented at the Fourth International Conference in Intelligent Tutoring Systems, San Antonio, TX, August 16-19, 1998.
4 Jonassen DH, Grabowski B. Handbook of Individual Differences, Learning, and Instruction. Hillsdale, NJ: Lawrence Erlbaum Associates, 1993.
5 Friedman R. Top ten reasons the World Wide Web may fail to change medical education. Acad Med. 1996;71:979–81.
6 McKimm J, Jollie C, Cantillon P. ABC of learning and teaching in medicine: Web-based learning. BMJ. 2003;326:870–3.
7 Brown BL. Web-based training. ERIC digest no. 218. Columbus, OH: ERIC Clearinghouse on Adult, Career, and Vocational Education, 2000.
8 Chumley-Jones HS, Dobbie A, Alford CL. Web-based learning: sound educational method or hype? A review of the evaluation literature. Acad Med. 2002;77(10 suppl):S86–93.
9 Wong G, Greenhalgh T, Russell J, Boynton P, Toon P. Putting your course on the web: lessons from a case study and systematic literature review. Med Educ. 2003;37:1020–3.
10 Collison G, Elbaum B, Haavind S, Tinker R. Facilitating Online Learning: Effective Strategies for Moderators. Madison, WI: Atwood Publishing, 2000.
11 Graff M. Cognitive style and attitudes towards using online learning and assessment methods. EJEL: Electronic Journal of eLearning 〈〉. 2003;1(1):21–8.
12 Alur P, Fatima K, Joseph R. Medical teaching websites: do they reflect the learning paradigm? Med Teach. 2002;24:422–4.
13 Merrill MD. Instructional strategies and learning styles: which takes precedence? In: Reiser R, Dempsey JV (eds). Trends and Issues in Instructional Design and Technology. Upper Saddle River, NJ: Merrill/Prentice Hall, 2002.
14 Cook DA, Dupras DM. A practical guide to developing effective Web-based learning. J Gen Intern Med. 2004;19:698–707.
15 Chen SY, Paul RJ. Individual differences in web-based instruction: an overview. Br J Educ Technol. 2003;34:385–92.
16 Curry L. An organization of learning style theory and constructs. In: Curry L (ed). Learning Style in Continuing Medical Education. Ottawa: Canadian Medical Association, 1983:115–23.
17 Curry L. Individual differences in cognitive style, learning style, and instructional preference in medical education. In: Norman G, Van der Vleuten C, Newble D (eds). International Handbook of Research in Medical Education. Dordrecht: Kluwer Academic Publishers, 2002:263–76.
18 Riding R, Cheema I. Cognitive styles: an overview and integration. Educ Psychol. 1991;11:193–215.
19 Rayner S. Reconstructing style differences in thinking and learning: Profiling learning performance. In: Riding R, Rayner S (eds). International Perspectives on Individual Differences, Volume 1: Cognitive Styles. Stamford, CT: Ablex Publishing Corporation, 2000.
20 Riding R. Cognitive Styles Analysis: Research Applications. Birmingham, UK: Learning and Training Technology, 2000.
21 Chen SY, Macredie RD. Cognitive styles and hypermedia navigation: development of a learning model. J Am Soc Inf Sci Technol. 2002;53(1):3–15.
22 Riding R. Cognitive style: a strategic approach for advancement. In: Riding R, Rayner S (eds). International Perspectives on Individual Differences, Volume 1: Cognitive Styles. Stamford, CT: Ablex Publishing Corporation, 2000.
23 Felder RM, Soloman BA. Index of learning styles 〈〉. Accessed 10 December 2004.
24 Felder RM, Silverman LK. Learning and teaching styles in engineering education. Engr Educ. 1988;78:674–81.
25 Abouserie R, Moss D. Cognitive style, gender, attitude toward computer-assisted learning and academic achievement. Educ Stud. 1992;18:151–60.
26 Liu M, Reed WM. The relationship between the learning strategies and learning styles in a hypermedia environment. Paper presented at the Annual Conference of the Association for Educational Communications and Technology, Nashville, TN, February 16-20, 1994.
27 Ford N, Chen SY. Individual differences, hypermedia navigation, and learning: an empirical study. J Educ Multimedia Hypermedia. 2000;9:281–311.
28 Ford N, Chen SY. Matching/Mismatching revisited: an empirical study of learning and teaching styles. Br J Educ Technol. 2001;32:5–22.
29 Pillay H. An investigation of the effect of individual cognitive preferences on learning through computer-based instruction. Educ Psychol. 1998;18:171–82.
30 Graff M. Learning from Web-based instructional systems and cognitive style. Br J Educ Technol. 2003;34:407–18.
31 Jonassen DH, Wang SR. Acquiring structural knowledge from semantically structured hypertext. J Comput Based Instruct. 1993;20(1):1–8.
32 Hsu TE, Frederick FJ, Chung ML. Effects of learner cognitive styles and metacognitive tools on information acquisition paths and learning in hyperspace environments. Paper presented at the National Convention of the Association for Educational Communications and Technology, Nashville, TN, February 16-20, 1994.
33 Triantafillou E, Pomportsis A, Demetriadis S, Georgiadou E. The value of adaptivity based on cognitive style: an empirical study. Br J Educ Technol. 2004;35:95–106.
34 Workman M. Performance and perceived effectiveness in computer-based and computer-aided education: do cognitive styles make a difference? Comput Hum Behav. 2004;20:517–34.
35 Parkinson A, Redmond JA. The impact of cognitive types and educational computer environments on learning performance. Dublin, Ireland: Trinity College, 2002.
36 Summerville J. Role of awareness of cognitive style in hypermedia. Int J Educ Technol. 1999 〈〉. Accessed 18 March 2004. [No longer available.]
37 Reed WM, Oughton JM, Ayersman DJ, Ervin JR Jr, Giessler SF. Computer experience, learning style, and hypermedia navigation. Comput Hum Behav. 2000;16:609–28.
38 Riding RJ, Rayner S. The information superhighway and individualized learning. Educ Psychol. 1995;15:365–78.
39 Felder RM. Reaching the second tier: learning and teaching styles in college science education. J Coll Sci Teach. 1993;23:286–90.
40 Kirby J, Moore P, Schofield N. Verbal and visual learning styles. Contemp Educ Psychol. 1988;13:169–84.
41 Effken JA, Doyle M. Interface design and cognitive style in learning an instructional computer simulation. Comput Nurs. 2001;19(4):164–71.
42 Montgomery SM. Addressing diverse learning styles through the use of multimedia. Paper presented at the 1995 ASEE/IEEE Frontiers in Education Conference, Atlanta, GA, November, 1995.
43 Kini AS. Effects of cognitive style and verbal and visual presentation modes on concept learning in CBI. Paper presented at the Annual Meeting of the American Educational Research Association, New Orleans, LA, April 4-8, 1994.
44 John D, Boucouvalas AC. The effect of cognitive style on user performance in tasks using different types of media. Paper presented at the EuroMedia 2001, Valencia, Spain, April, 2001.
45 Cunningham-Atkins H, Powell N, Moore D, Hobbs D, Sharpe S. The role of cognitive style in educational computer conferencing. Br J Educ Technol. 2004;35:69–80.
46 Kolb D. Experiential Learning: Experience as the Source of Learning and Development. Englewood Cliffs, NJ: Prentice-Hall, 1984.
47 Kolb D. Learning Style Inventory, Version 3. Boston, MA: Hay Group, 1999.
48 Ruble TL, Stout DE. Learning styles and end-user training: an unwarranted leap of faith. MIS Q. 1993;17:115–8.
49 Garner I. Problems and inconsistencies with Kolb’s learning styles. Educ Psychol. 2000;20:341–9.
50 Webb MW. A definitive critique of experiential learning theory 〈∼mnwebb/critique.htm〉. Accessed 10 December 2004.
51 Bostrum RP, Olfman L, Sein MK. Learning styles and end-user training: a first step. MIS Q. 1993;17:118–20.
52 Kolb D, Boyatzis R, Mainemelis C. Experiential learning theory: previous research and new directions. In: Sternberg R, Zhang L (eds). Perspectives on Thinking, Learning, and Cognitive Styles. Mahwah, NJ: Lawrence Erlbaum Associates, 2001.
53 Rezler AG, Rezmovic V. The learning preference inventory. J Allied Health. 1981;10:28–34.
54 Jung CG. Psychological Types. Princeton, NJ: Princeton University Press, 1976.
55 Briggs Myers I, Briggs KC. Myers-Briggs Type Indicator 〈〉. Accessed 10 December 2004. CPP, Inc., Palo Alto, CA.
56 Kirchhoff KT, Holzemer WL. Student learning and a computer-assisted instructional program. J Nurs Educ. 1979;18(3):22–30.
57 Thompson EC. Computer-assisted instruction in curricula of physical therapist assistants. Phys Ther. 1987;67:1237–9.
58 Brudenell I, Carpenter CS. Adult learning styles and attitudes toward computer assisted instruction. J Nurs Educ. 1990;29:79–83.
59 Lowdermilk DL, Hopkins Fishel A. Computer simulations as a measure of nursing students’ decision-making skills. J Nurs Educ. 1991;30:34–9.
60 DiBartola LM, Miller MK, Turley CL. Do learning style and learning environment affect learning outcome? J Allied Health. 2001;30:112–5.
61 Engleberg NC, Schwenk T, Gruppen LD. Learning styles and perceptions of the value of various learning modalities before and after a 2nd-year course in microbiology and infectious diseases. Teach Learn Med. 2001;13:253–7.
62 Lynch TG, Steele DJ, Johnson Palensky JE, Lacy NL, Duffy SW. Learning preferences, computer attitudes, and test performance with computer-aided instruction. Am J Surg. 2001;181:368–71.
63 Steele DJ, Johnson Palensky JE, Lynch TG, Lacy NL, Duffy SW. Learning preferences, computer attitudes, and student evaluation of computerised instruction. Med Educ. 2002;36:225–32.
64 Fleming DE, Mauriello SM, McKaig RG, Ludlow JB. A comparison of slide/audiotape and Web-based instructional formats for teaching normal intraoral radiographic anatomy. J Dent Hyg. 2003;77:27–35.
65 Bostrum RP, Olfman L, Sein MK. The importance of learning style in end-user training. MIS Q. 1990;14:101–19.
66 Hart G. Learning styles and hypertext: exploring user attitudes. Paper presented at the Australian Society for Computers in Learning in Tertiary Education, Melbourne, December 3-7, 1995.
67 Rourke L, Lysynchuk L. The influence of learning style on achievement in hypertext. Paper presented at the annual meeting of the American Education Research Association, New Orleans, LA, April 24-28, 2000.
68 Melara GE. Investigating learning styles on different hypertext environments: hierarchical-like and network-like structures. J Educ Comput Res. 1996;14:313–28.
69 Clariana RB. Considering learning style in computer-assisted learning. Br J Educ Technol. 1997;28:66–8.
70 Yoder ME. Preferred learning style and educational technology: linear vs interactive video. Nurs Health Care. 1994;15:128–32.
71 Liegle JO, Janicki TN. Effect of learning styles on the navigational needs of computer-based training module learners. Paper presented at the ISECON 2001, Cincinnati, OH, November 1-4, 2001. Vol. 18.
72 Jordanov WL. An examination of the relationship between learning style and technology use. Paper presented at the Annual Meeting of the Mid-South Educational Research Association, Little Rock, AR, November, 2001.
73 Clark R. Confounding in educational computing research. J Educ Comput Res. 1985;1:28–42.
74 Friedman C. The research we should be doing. Acad Med. 1994;69:455–7.
75 Cook D. The research we still are not doing: An agenda for research in computer-based learning. Acad Med. [in press].
76 Billings DM, Cobb KL. Effects of learning style preferences, attitude and GPA on learner achievement using computer assisted interactive videodisc instruction. J Comput Based Instruct. 1992;19:12–6.
    77 Allen LR. Measuring attitude toward computer assisted instruction: the development of a semantic differential tool. Comput Nurs. 1986;4:144–51.
      78 Lehmann HP, Freedman JA, Massad J, Dintzis RZ. An ethnographic, controlled study of the use of a computer-based histology atlas during a laboratory course. J Am Med Inform Assoc. 1999;6:38–52.
        79 Lieberman G, Abramson R, Volkan K, McArdle PJ. Tutor versus computer: a prospective comparison of interactive tutorial and computer-assisted instruction in radiology education. Acad Radiol. 2002;9:40–9.
          80 Miller JA. Enhancement of achievement and attitudes through individualized learning-style presentations of two allied health courses. J Allied Health. 1998;27:150–6.
            81 Bland CJ, Meurer LN, Maldonado G. A systematic approach to conducting a non-statistical meta-analysis of research literature. Acad Med. 1995;70:642–53.
              82 Oltman P, Raskin E, Witkin H. Group Embedded Figures Test 〈〉. Accessed 10 December 2004. Mind Garden, Inc., Redwood City, CA.
                83 Dunn R, Dunn K, Price G. Productivity Environmental Preference Survey. Lawrence, Kansas: Price Systems, 1996.
                  84 Aragon SR, Johnson SD, Shaik N. The influence of learning style preferences on studentsuccess in online vs. face-to-face environments. Paper presented at the WebNet 2000 World Conference on the WWW and Inter-net, San Antonio, TX, October 30-November 4, 2000.
                    85 Ghinea G, Chen SY. The impact of cognitive styles on perceptual distributed multimedia quality. Br J Educ Technol. 2003;34:393–406.
                      86 Packard AL, Holmes GA, Viveiros DM, Fortune JC. Incorporating individual learning styles into computer assisted environments for instruction basic research and statistical concepts. Paper presented at the Eastern Ed-ucational Research Association, Hilton Head, SC, February, 1997.
                        87 Shih C-C, Gamon J. Web-based learning: relationships among student motivation, attitude, learning styles, and achievement. J Agric Educ. 2001;42(4):12–20.
                          88 Shih C-C, Gamon JA. Relationships among learning strategies, patterns, styles, and achievement in Web-based courses. J Agric Educ. 2002;43(4):1–11.
                            89 Sternberg RJ. Thinking Styles. New York: Cambridge University Press, 1997.

                              *Discussing a learning style along with a cognitive style may seem inappropriate. Extrapolating the learning style (a more narrowly defined and less stable construct) to make inferences about the cognitive style (which is more stable and holds implications for both learning and nonlearning contexts) is certainly inappropriate. However, in a specific learning setting a learner’s cognitive style will likely have effects similar to an analogous learning style. Thus, when a cognitive style and learning style have congruent definitions, and both are applied to a learning context, I will discuss the two as a single construct.
                              Cited Here...

                              © 2005 Association of American Medical Colleges