Secondary Logo

Journal Logo

Medical Education

Viewpoint: Introducing the Concept of Epistemological Beliefs into Medical Education: The Hot-Air-Balloon Metaphor

Roex, Ann, MD; Degryse, Jan, MD, PhD

Author Information
doi: 10.1097/ACM.0b013e3180556abd
  • Free


Over the last decade, medical educators have worked extensively to translate the findings of medical expertise research into practical consequences for medical education. As a result, most medical curricula now include foci on clinical knowledge, skills, reasoning processes, and critical analysis. There is a general agreement that an intense interaction between self-analysis and action enhances the learning process. More attention has therefore been paid to metacognitive skills, self-efficacy, self-evaluation, and self-reflection. This has not happened without controversy (because of the low observed correlations between the findings from self-assessment and from assessment by external criteria, and because of the methodological problems in studying these correlations).1

Although much energy has been spent on revolutionary curriculum changes in medical schools, some more recent developments in educational psychology, such as insights into students' epistemological beliefs, have yet to find their way into the curriculum. We wrote this article to introduce the concept of epistemological beliefs in the context of existing models of medical expertise. The most relevant question for medical education is whether the introduction of the dimension of epistemological beliefs in existing instructional environments (e.g., by training students to articulate their epistemological beliefs or by confronting students with reasoning processes based on more sophisticated levels of epistemological beliefs) would enhance students' medical competence and performance.

What Are Epistemological Beliefs?

As with many other research topics, different names and different definitions have been used for the study of what we, in this article, are calling epistemological beliefs. Those names include epistemological theories, ways of knowing, epistemic reflection, epistemological resources, epistemic cognition, and epistemological thinking, to mention a few. However, they all refer to the cognitions (i.e., understandings) an individual (for our purposes in this discussion, a student) has about knowledge and knowing.2 The different theories on epistemological beliefs differ to the extent to which they also include a student's cognitions about learning, intelligence, instruction, etc.2

The research into epistemological beliefs was initiated by Perry in the 1970s and has since been approached in three different ways. A first group of researchers saw personal epistemology as a one-dimensional, generic characteristic with a step-by-step developmental progression over time, ranging from a dualistic view of knowledge (i.e., knowledge is right or wrong and absolutely observable) to a relativistic view of knowledge (i.e., knowledge is actively constructed by the knower using context-dependent evidence).3 Within this framework, semistructured interviews have most frequently been used to assess the research participant's personal epistemology. The developmental pattern is well described and chronological. Changes are said to be induced by confrontations with problems that need more sophisticated cognitive processing to solve.

This one-dimensional construct was challenged by Schommer.4 She introduced a second framework, proposing the existence of different dimensions—epistemological beliefs—that develop independently of one another. In this approach, she used a questionnaire to assess the different beliefs. The easy use and applicability of this questionnaire has guided many other researchers along the same pathway. Several of these researchers have formulated the same four epistemological beliefs: the belief in

  • the certainty of knowledge (varying from the belief that knowledge is certain, to the belief that knowledge is tentative and evolving),
  • the simplicity of knowledge (from “knowledge is simple” to “knowledge is an accumulation of interrelated facts”),
  • the source of knowledge (from “knowledge resides outside of oneself and is transmitted passively” to “knowledge is actively constructed by the knower”), and
  • the justification of knowing (based either on observation or on authority, or, instead, as a result of inquiry and active processing by the knower).

Initially, epistemological beliefs were seen as generic—that is, not dependent on any specific domain of knowledge. However, several researchers have demonstrated domain specificity.5 Questionnaires have permitted researchers to study the relationship between epistemological beliefs and age or education, choice of majors, and learning outcomes such as strategy use, comprehension, text processing, and motivation. (For a detailed overview, see Buehl, MM. At the Crossroads of Epistemology and Motivation: Modeling the Relations between Students' Domain-Specific Epistemological Beliefs, Achievement Motivation, and Task Performance [unpublished dissertation]. University of Maryland; 2003).

Within this second framework, Hofer has highlighted the metacognitive aspect of epistemological beliefs:

For example, as you build your own knowledge about a topic that is previously unfamiliar, how will you evaluate and assess the veracity of what you read and hear? Whose authority will you accept—and why? What evidence will you decide is acceptable justification for particular recommended choices of action? How certain are you that what you read is true, that it is supportable, that it can be believed? How will you reconcile your own experience with those of experts? When do you decide that you know enough and that your understanding is adequate?6

The strength of this approach is that it applies not only to ill-structured or complex problems (these are problems for which no single right or wrong answers are available) but also to everyday confrontations with (new) information. Moreover, it offers a theoretical framework for understanding and studying individual approaches to the vast supply of information.

Recently, Hammer and Elby7 and Louca et al8 have proposed an innovative approach to personal epistemology. They claim that any individual possesses a wide range of epistemological “resources,” which are seen as “grain-sized” cognitive elements for understanding and approaching knowledge. Depending on the nature of the presented stimulus and its context, a certain number of resources will be activated. This model explains observed variations in the epistemological beliefs that are manifested in individuals' reasoning processes. Growth of personal epistemology is attributable to an “increasing interconnection, compilation and articulation of resources.”7,8

The differences between all the different lines of research on the nature of personal epistemology are attributable to different assumptions about the topic, which have prompted different research methods, which in turn have fostered different study foci and results. First is the evolution beyond viewing epistemological beliefs as being domain general (King and Kitchener, Schommer), domain specific (Hofer, Buehl), or problem- and context specific (Louca). Second, epistemological beliefs have been considered to be one dimensional (King and Kitchener) or multidimensional (Schommer, Hofer, etc.).

Possible Role of Epistemological Beliefs in Medicine

There are several reasons why the study of epistemological beliefs could be highly relevant to medical education.

First, the exponential growth in the number and accessibility of information sources increases the relevance of Hofer's above-cited questions regarding the acquisition of knowledge. Second, epistemological beliefs typically come to the surface when one is confronted with ill-structured problems, for which right or wrong answers are not available.9 In medicine, the occurrence of ill-structured problems is very high, and straightforward answers are rarely available.10,11 Third, while emphasizing the use of EBM in medical practice, equal attention needs to be given to how such (scientific) knowledge is to be understood, integrated with existing knowledge, and applied to individual patients. Scientific knowledge has to be understood from the perspective in which it was found, and it has to be applied correctly in each specific context. For example, a physician may have a set of treatment guidelines that reflect a compromise between available evidence and cost-benefit analysis (at the population level) but that are also influenced by epidemiological, social, cultural, and political factors.12,13 A physician who takes the guideline's origin and context into account will understand the guidelines more thoroughly and will be able to apply them better to his or her own patient.

Possible Role of Epistemological Beliefs in Medical Expertise

The medical expertise literature is vast and has varied focal points, but we have found no or very little reference to epistemological beliefs, though some authors draw attention to the necessity of having ill-structured problems in medical instructional programs and assessments in teaching clinical competence.13 We will use a metaphor of a piloted hot-air balloon in an attempt to illustrate the need for introducing epistemological beliefs in medical expertise models (see Figure 1). This metaphor symbolizes the dynamic nature of expertise, which has to be seen rather as a state than as a trait.14,15

Figure 1
Figure 1:
The hot-air-balloon metaphor symbolizes the dynamic nature of expertise. To become an expert clinician, one first must have a strong, well-woven wicker basket (knowledge base), a well-designed envelope (of which each of the gores represents a different skill), and a well-equipped burner (motivation, intelligence, and other noncognitive factors). Second, all these items need to be present in the right proportions and in the right positions. A small envelope will not lift a big, heavy basket from the ground, and an excessively strong burner could ignite the envelope. Third, a pilot who reflects on his or her own actions and thinking (metacognitive abilities) and who is aware of and uses the external forces on the balloon (epistemological beliefs) will inspire more confidence in the balloon's passengers. Balloon trips are possible from the moment that a minimal basket, envelope, and burner are available and in logical proportions to each other. For as long as the pilot is self-reflecting, increasing flights will strengthen the balloon's structures and will thus allow the pilot to transport more passengers or to travel in more intricate situations. Recognition and acceptance of the changeability of the external forces on the balloon (wind, weather, and landscape features) will allow the pilot to use these to navigate the balloon. Medical curricula and assessment procedures should bear in mind the equivalents of the balloon's different parts to design curricula that will produce physicians with high chances of expertise acquisition and to expand the scope of existing medical curricula beyond the traditional foci of extension of knowledge bases (classical curricula) or problem-solving skills (PBL).

Although theories on the organization of knowledge diverge (e.g., script theory, instance-based frameworks, prototypes, or semantic networks), recent research suggests that a single individual may employ different methods of knowledge storage.11,16–18 Such research confirms that a well-structured “basket” of knowledge provides an essential basis for acquiring medical expertise. The hot-air balloon's wicker basket symbolizes the need for such an organized knowledge base.

The hot-air-balloon's envelope (i.e., air bag) represents the range of skills that are required (varying from technical skills to interpersonal skills and learning skills, and, last but not least, reasoning skills). The reasoning processes (e.g., pattern recognition versus reasoning at the pathophysiological level) an individual uses are influenced by the way his or her knowledge base is organized in that particular domain (e.g., storage of previous encounters versus storage of pathophysiological knowledge). This relationship is symbolized by the balloon's cords, which tie the envelope to the basket. Motivation, intelligence, and other noncognitive factors (the balloon's burners) provide the necessary energy for undertaking balloon trips.19 The role of the burners is clearly crucial, but it is equally important that they are activated at the right speed and at the right stages.

Metacognition (initially defined by Flavell20 as “one's knowledge on one's own cognitive processes and all the products or anything related to them”) has been the focus of growing attention in both medical curricula and in medical expertise models.11,21 This trend has not been without controversy. First, students' capacity to self-reflect has—not withstanding methodological problems—repeatedly proven to be low. Second, no consensus has yet been reached on the mechanism of self-reflection: is metacognition a generic ability, or does it become possible only after—through experience—certain tasks have become automized? (Once certain tasks are automized, physicians need to spend less cognitive activity to be able to resolve the task, and thus cognitive space becomes available for metacognitive activity.) The pilot–physician's metacognitive abilities play an important, but not determinate, role: they raise his or her awareness of the basket's weak points, and they allow the pilot–physician to transport the balloon's passengers by placing them at the strong side of the basket. Pilots who reflect on the skills used to maneuver the hot-air balloon can learn faster from their acts and might thus be able, much earlier in training, to predict the consequences of different actions.

The possession of a well-constructed wicker basket (knowledge), a well-balanced envelope (skills), powerful burners (noncognitive factors), and self-reflective skills (metacognition) are essential to a successful balloon trip. However, two pilots with similar equipment will not necessarily navigate with the same ease or instill the same confidence in their passengers. All pilots have to be aware of the existence of external forces on the balloon (wind, weather). But the well-equipped pilot only becomes an expert pilot when he or she accepts and even uses the changeability of the external forces (e.g., the different wind currents depending on altitude) to navigate the balloon safely towards its destination. The theories of epistemological beliefs study the extent to which pilots believe in the relativistic nature and context dependency of navigation rules. In a parallel way, epistemological beliefs can play a role in medical expertise and, because of their potential importance, are worthy of investigation.


The question of what makes one an expert in a field has fascinated many researchers since the 1960s. Departing from different points of interest (cognitive psychology, artificial intelligence, medical clinical reasoning, and others), several lines of research have developed their own explanatory theories. Researchers currently agree on the complexity of the phenomenon of medical expertise, and the vast array of theories makes it difficult to give an overview on this domain of study. The introduction of the concept of epistemological beliefs in medical expertise models is appealing because it offers a framework for phrasing and studying issues that intuitively seem essential to educators, such as the need for ill-structured questions in medical curricula and assessments10 and the need for students to develop skills to correctly interpret and apply scientific insights to daily practice.11

The hot-air balloon metaphor is not intended to be a new model for medical expertise acquisition. It is rather a way of effectively illustrating the different factors contributing to medical expertise (based on existing literature) in a logical and surveyable way, and it points out the need to include epistemological beliefs into medical expertise theories. Because of its visual impact, the hot-air-balloon metaphor provides a facilitative framework for designing medical curricula, assessment procedures, and educational research projects. Medical curricula and assessment procedures should bear in mind the equivalents of the balloon's different parts to design curricula that will produce physicians with high chances of expertise acquisition and to expand the scope of existing medical curricula beyond the traditional foci of extension of knowledge bases (classical curricula) or problem-solving skills (PBL).

Deciding which of the frameworks for epistemological beliefs is most suitable in the medical context is a complex issue, but to investigate its importance, a framework must be chosen.2,22 Niessen et al22 criticize the absence of clear description of underlying default assumptions in many of the existing theories. They propose a dimensional framework that offers guidance to researchers to define these aspects and that allows them to map out the differences between the existing approaches to epistemological beliefs.

Physicians involved in medical education are familiar with the phenomenon of case specificity.23 Many among them might also have observed that students who are successful in critically analyzing a case in a certain situation (e.g., an EBM course) are not always able to transfer these critical reasoning processes to a similar case in another setting (e.g., in practice learning). As such, the framework of epistemological resources (see above) might sound more appropriate to many medical educators. Unfortunately, adopting the resources theory raises many obstacles to the study of epistemological beliefs. First, when, where, and how can we reliably measure the epistemological resources if they are so highly context dependent? Second, the context dependency and variety in use of epistemological resources implies the use of different measures; how to combine the results of these different measures? Thirdly, how can one define change? The absence of clear-cut answers to these questions reduces the applicability of this framework and probably explains the reticence many have shown to epistemological beliefs research.

Hofer's metacognitive approach to epistemological beliefs is very appealing in the current medical context of a rapidly changing stream of information, and in our current medical culture, in which doctors are continually challenged to correctly interpret, use, and apply guidelines and existing scientific data, it offers a framework for naming these aspects that we intuitively find important and that deserve more attention than they have received up to now.

Within the literature on epistemological beliefs, the different beliefs (such as the belief in the certainty of knowledge, in the structure of knowledge, etc.) and their relationships to various other outcomes (such as academic achievement, motivation, learning styles, strategy use) have been extensively studied. Further research will have to establish whether the same beliefs are relevant in the medical context and, if so, how these beliefs relate to the different factors in the medical expertise model and to other outcome measures.

It is, for example, likely that the belief in the structure of knowledge influences how the knowledge base is organized: the belief that knowledge is simple could be reflected in a compartmental organization of the knowledge base, which, in the balloon analogy, would lead to a weaker wicker basket than when knowledge is assumed to be complex, which would be shown by a tightly woven basket. In her research into the link between students' belief in the source of knowledge and their motivation, Buehl has demonstrated that the belief in oneself as a constructor of knowledge leads to a higher level of motivation in history students.

Researchers have repeatedly described the discrepancies between the epistemological beliefs individuals claim to have (professed beliefs) and the beliefs that are expressed through their behaviors (enacted beliefs). Other researchers have warned of the influence of social desirability on students' descriptions of their epistemological beliefs in questionnaires or interviews. Both these phenomena dictate the use of—at least where possible—measures “in real time,” and both must be taken into account during the design of studies.

Until now, in their studies of epistemological beliefs, educational psychologists have used a diversity of outcome measures: learning strategies, learning styles, critical thinking, etc. The main concern of medical educators is how to more efficiently train students to gain medical expertise. The challenge of including this concern in the study of medical students' epistemological beliefs will only be successful if it is reflected in and related to relevant outcome measures for medical education. After all, the core question is: should we pay more attention to epistemological beliefs in learning environments and assessment methods, to train more competent physicians?

Future Research

In this article, we have tried to show that the concept of epistemological beliefs provides a valuable addition to medical expertise theories and potentially makes medical expertise theories more applicable to medical education. The hot-air-balloon metaphor provides a tool that allows researchers and educators to reflect on and advance insights about epistemological beliefs into medical curricula and into educational research. The first items on the research agenda are, on the one hand, to investigate the link between physicians' epistemological beliefs and their competence and, on the other, to explore the influence of instructional design on the development of physicians' epistemological beliefs. To successfully answer these questions, researchers have to clearly define the concept of epistemological beliefs, its scope, and an underlying framework, and then develop appropriate measures to map out the role of epistemological beliefs in a realistic medical setting.


The authors are grateful to Professor Georges Bordage and to the anonymous reviewers for their helpful comments and suggestions.


1 Ward M, Gruppen LD, Regehr G. Measuring self-assessment: current state of the art. Adv Health Sci Educ Theory Pract. 2002;7:63–80.
2 Pintrich PR. Future challenges and directions for theory and research on personal epistemology. In: Hofer BK, Pintrich PR, eds. Personal Epistemology: The Psychology of Beliefs about Knowledge and Knowing. Mahwah, NJ: Laurence Erlbaum; 2002:389–413.
3 King PM, Kitchener KS. Developing Reflective Judgement. New York, NY: Jossey-Bass; 1994.
4 Schommer M. Effects of beliefs about the nature of knowledge on comprehension. J Educ Psychol. 1990;82:498–504.
5 Hofer BK. Dimensionality and disciplinary differences in personal epistemology. Contemp Educ Psychol. 2000;25:378–405.
6 Hofer BK. Epistemological understanding as a metacognitive process: thinking aloud during online searching. Educ Psychol. 2004;39:43–55.
7 Hammer D, Elby A. On the form of a personal epistemology. In Hofer BK, Pintrich PR, eds. Personal Epistemology: The Psychology of Beliefs about Knowledge and Knowing. Mahwah, NJ: Laurence Erlbaum; 2002:169–190.
8 Louca L, Elby A, Hammer D, Kagey T. Epistemological resources: applying a new epistemological framework to science instruction. Educ Psychol. 2004;39:57–68.
9 Kitchener KS. Cognition, metacognition, and epistemic cognition—a 3-level model of cognitive processing. Hum Dev. 1983;26:222–232.
10 Elstein AS, Kleinmuntz B, Rabinowitz M, et al. Diagnostic reasoning of high- and low-domain-knowledge clinicians: a reanalysis. Med Decis Making. 1993;13:21–29.
11 Gruppen LD. Pre-Conference Paper No. 1—Seventh Cambridge Conference “Moving Medical Education from the Hospital to the Community.” Cognitive and Educational Psychology Foundations for Ambulatory-Based Education. Ann Arbor, MI. July 8–15, 1995;93–117.
12 Asch DA, Hershey JC. Why some health policies don't make sense at the bedside. Ann Intern Med. 1995;122:846–850.
13 Elstein AS. Beyond multiple-choice questions and essays: the need for a new way to assess clinical competence. Acad Med. 1993;68:244–249.
14 Eraut M. Expert and epertise: meanings and perspectives. Learn Health Soc Care. 2005;4:173–179.
15 Van der Vleuten CPM. The assessment of professional competence: developments, research and practical implications. Adv Health Sci Educ. 1996;41–67.
16 Custers EJFM, Regehr G, Norman GR. Mental representations of medical diagnostic knowledge: a review. Acad Med. 1996;71:555–561.
17 Elstein AS, Schwarz A. Clinical problem solving and diagnostic decision making: selective review of the cognitive literature. BMJ. 2002;324:729–732.
18 Schmidt HG, Norman GR, Boshuizen HPA. A cognitive perspective on medical expertise: theory and implications. Acad Med. 1990;65:611–621.
19 Alexander PA. Mapping the multidimensionsal nature of domain learning: the interplay of cognitive, motivational and strategic forces. Adv Motiv Achievement. 1997;10:213–250.
20 Flavell JH. Metacognitive aspects of problem solving. In: Resnick LB, ed. The Nature of Intelligence. Hillsdale, NJ: Lawrence Erlbaum Associates; 1976.
21 Dunphy BC, Williamson SL. In pursuit of expertise. Adv Health Sci Educ. 2004;9:107–127.
22 Niessen T, Vermunt JD, Abma T, Widdershoven G, van der Vleuten C. On the nature and form of epistemologies: revealing hidden assumptions through an anysis of instrument design. Eur J School Psychol. 2004;2:39–64.
23 Elstein AS, Shulman LS, Sprafka SA. Medical Problem Solving: An Analysis of Clinical Reasoning. Cambridge, Mass: Harvard University Press; 1978.
© 2007 Association of American Medical Colleges