Chapter 3: Uncertainty Tolerance Conceptual Modelling

Research into the uncertainty tolerance construct has occurred since the mid-20th century. It started in the field of personality research, findings from which were drawn largely from studies of children and undergraduate students (Budner, 1962; Frenkel-Brunswik, 1949) and tended to position uncertainty tolerance as a personality trait. Historically, how an individual perceives and responds to uncertainty was considered largely immutable and unchanged by factors such as learning or experience (Budner, 1962; Hillen et al., 2017). Contemporary research is changing the dominance of this perspective (Patel et al., 2022; Stephens et al., 2021, 2022b; Strout et al., 2018). There is increasing evidence, including from studies with health professions learners, that uncertainty tolerance has features that are state specific, i.e., context dependent, changeable, and contingent (Hillen et al., 2017; Lazarus et al., 2022; Stephens et al., 2022b).

Research findings increasingly indicate that the ways in which educators develop and execute curricula can influence the ways in which learners understand and respond to uncertainty (Lazarus et al., 2022; Patel et al., 2022; Stephens et al., 2022b). The goal of this handbook is to help health professions educators consider how teaching practices and curricula design can support learners’ adaptive, as opposed to maladaptive, responses to uncertainty, in preparation for managing uncertainty in their future healthcare careers. As a first step towards understanding how theory about the uncertainty tolerance construct can inform educational practice, this chapter outlines a conceptual model for uncertainty tolerance.

Conceptual Modelling of Uncertainty Tolerance

No model is perfect. Considering the complexity of the uncertainty tolerance construct, this imperfect nature of modelling holds true. Uncertainty tolerance appears to be influenced by both intrinsic factors (e.g., personality traits, such as introversion) and extrinsic factors (e.g., the context in which the uncertainty is experienced); therefore, a single model is unlikely to capture all the nuances of the construct or a complete picture of how it manifests in the real world. As Hillen et al. (2017) point out, however, models of uncertainty tolerance do help to explain different aspects of the construct and may be engaged to inform education and drive research; existing models can serve as frameworks for considering the relationship between healthcare and educational environments and uncertainty tolerance development.

The integrative model of uncertainty tolerance developed by Hillen et al. (2017) (opens in new tab) is a useful starting point. Its advantages are the extensive literature base upon which it is built and the psychological lens applied to it. For this reason, several studies in health professions education engage this model as part of data analysis (Stephens et al., 2021, 2022a, 2022b, 2023; van Iersel et al., 2019). Hillen et al.’s (2017) model consists of the following dimensions: stimulus, perception, moderators, and response; a structure which is consistent with many psychological constructs (Pearce, 1987; Ahmed, 2016). These are introduced in Video 1.1 and explained in greater detail in the sections that follow.

Model of Uncertainty Tolerance for Health Professions’ learners

This section describes a conceptual model of uncertainty tolerance based on these psychological “stimulus-response” models, as well as research within the health professions educational context (Stephens, 2022; Stephens et al., 2022b, 2023). This model of uncertainty tolerance for health professions learners differs from the Hillen et al.’s (2017) integrative model of uncertainty tolerance (described in Video 1.1) by the addition of a feedback loop, which recognises changes in perceptions of and responses to uncertainty. For example, critical reflection and the outcomes of prior uncertain scenarios have been shown to influence future perceptions of and responses to uncertainty (Stephens et al., 2022b).

 

Stimulus

Uncertainty stimuli are phenomena or aspects of an event which elicit a perception that something is unknown (Han et al., 2011; Hillen et al., 2017). As described in Chapter 1, a stimulus may be experienced as uncertain (i.e., the source of uncertainty) because of probability (inability to predict the future), ambiguity (incomplete, inconsistent, or unreliable information), or complexity (the numerous parts of a phenomena being so intricately tied together that it is near impossible to tease them apart). One or more stimuli may lead to perceptions of uncertainty. Click through the slides below as you consider Case Study 3.1a.

Chest pain is a common and important emergency presentation, and hence a common topic in the education and practice of many health professionals. Although the hospital Zach is visiting is likely to have protocols for evaluating patients with chest pain, there are elements in Zach’s case that, at least with the present information, represent ambiguity, complexity, and probability.

Zach is unable to provide definitive answers to all of Amir’s (the triage nurse’s) questions. For example, whether any family members have had a ‘heart attack’ is unknown. This is representative of incomplete information, an example of ambiguity. At this stage, the reason why Zach is experiencing chest pain is unknown, but there are several diagnostic clues. Zach describes the chest pain as ‘crushing’, which could suggest an acute myocardial infarction, or heart attack. However, Zach has no known risk factors for heart disease. The fact that Zach experiences mild anxiety could raise a panic attack as an alternative diagnostic cause of his chest pain. There are also many other causes of chest pain that are yet to be explored in this encounter. These multiple cues for interpretation, which may be interdependent, add complexity to Zach’s case. Furthermore, it isn’t possible to fully predict what will happen next to Zach, which represents probability as another stimulus of uncertainty in the case.

Perceptions and Moderators

If there are properties of information that could stimulate uncertainty, but there isn’t anyone around to perceive these, does uncertainty exist? Hillen et al.’s (2017) integrative model of uncertainty tolerance suggests that without the human perception that uncertainty is present, uncertainty doesn’t exist. Imagine you are Amir, the triage nurse, who assessed Zach. Would you perceive uncertainty? Whether you would or not, and how you would respond, are likely to be influenced by a range of different moderators. Click through the slides below and consider Scenarios A & B which extend from Case Study 3.1a.

In Scenario A, you might feel some uncertainty. This scenario illustrates how other factors related to experience, such as having a clear sense of the scope of your role, knowledge of the available resources, and awareness of how to navigate the local healthcare context, all facilitate whether you perceive uncertainty and, if you do, your capacity to adaptively respond to the uncertainty in Zach’s presentation and effectively care for him. In Scenario B, your knowledge of and experience with managing chest pain might be just as comprehensive; however, different experiential factors, such as your recent experience learning about a bad outcome for a person with a similar presentation and uncertainty about your role in the healthcare team, could influence your perceptions of the uncertainty in Zach’s case.

An individual’s perceptions of and responses to uncertainty are impacted by a variety of moderators (Figure 3.1) (Hillen et al., 2017; Stephens et al., 2022b). Moderators encompass individual, sociocultural, and environmental factors and may result in an individual having more capacity to effectively manage (i.e., adaptively respond to) uncertainty or having less capacity to effectively manage (i.e., maladaptively respond to) uncertainty. Importantly, the moderator of perceived stakes (e.g., the likelihood that a bad outcome could occur, and the severity of that outcome) can outcompete other moderators (Gupta & Fogarty, 1993; Makkawi & Rutledge, 2000) in moderating an individual’s uncertainty tolerance. This makes sense, particularly in healthcare settings, where the stakes include life-altering or life-ending outcomes for patients and professional ramifications for healthcare professionals deemed to have contributed to such outcomes.

The sustainability literature – in which the relationship between complexity, stakes, and uncertainty is often discussed – positions stimuli of uncertainties along a spectrum (Diwekar et al., 2021; Funtowicz, 2020; Lazarus & Funtowicz, 2023; Williams, 2012). At one end of the spectrum is quantifiable uncertainty (risk), and at the other end is irreducible uncertainty (complexity). The decision-making process changes depending on where along the uncertainty continuum a problem sits and how high the stakes are in the given context. Therefore, what is defined as an adaptive or maladaptive response will depend on whether the uncertainty stimulus is risk or complexity.

Translating this continuum to healthcare, when risks or stakes are low (e.g., treating a urinary tract infection in an otherwise healthy, cis-gendered female adult), an adaptive response to uncertainty includes evidence-based approaches (e.g., using the first-line antibiotic in that setting). However, if the stakes and/or the complexity are high (e.g., irreducible uncertainty), an adaptive response to uncertainty may include expanding sources of expertise beyond the existing evidence base, considering the lived experience of the person seeking care, or involving other healthcare professionals, such as those in different professions or specialties. Discussions within this ‘extended peer community’ (Funtowicz & Ravetz, 1994) focus on identifying the next best step in treating an individual as opposed to a definitive diagnosis. For example, consider the context of an older person in hospital whose life is at risk and who has multiple health conditions, each of which, or the management thereof, may impact the others. Due to the complexity in this case, it isn’t clear which evidence and guidelines may be applied. In such cases, drawing on diverse expertise and acknowledging relevant values (e.g., of the patient, health professionals in the community who know the patient well, and so on) to identify the next step would be considered an adaptive response.

Experience is a moderator commonly discussed in the uncertainty tolerance literature (Ilgen et al., 2022; Lazarus et al., 2022; Stephens et al., 2021, 2022b; Strout et al., 2018). In both Scenario A & B extending from Case 3.1a, the nurse is described as ‘very experienced’. Gaining a breadth and depth of experience is an individual factor that may moderate one’s uncertainty tolerance towards adaptive responses (Stephens et al., 2022b). However, this is not always the case, as the nature or quality of experience, rather than simply the amount of experience, seems to influence how individuals perceive and respond to uncertainty (Stephens et al., 2022b).

Consider the potential traumatic psychological impacts of different experiences or uncertain events that result in bad outcomes. In Scenario B, the nurse’s recent experience of learning about a missed aortic dissection may negatively influence their uncertainty tolerance when caring for a person presenting with similar symptoms, due to the psychological distress of a prior experience. Another example of an experience that could negatively influence uncertainty tolerance is where a healthcare professional observes a rare but significant side effect of a medication or procedure.

These examples illustrate the potential impact of situational stakes superseding the impact of experience. The threat to human life influences how the healthcare professional may respond to the uncertainty, despite the presence of other moderating factors, such as experience, clear roles, and teamwork. Researchers are still learning how moderators interact with each other and how they influence people’s perceptions of and responses to uncertainty.

Responses

Hillen et al. (2017) divide the ways in which an individual responds to uncertainty into cognitive (thinking), emotional (feeling), and behavioural (acting) domains. They then place each domain on a continuum from positive to negative and provide descriptions of specific responses related to their respective positions at each end of this spectrum. For example, a negative response to uncertainty may involve thinking of the situation as threatening, feeling worried or fearful, avoiding the source of uncertainty, or deferring making decisions. By contrast, a positive response may include thinking of the situation as an opportunity, feeling calm and curious, approaching the source of uncertainty, and gathering more information.

Although these descriptions may serve as a helpful starting point to determine whether a response is adaptive or maladaptive, a major limitation in trying to divide responses into positive and negative categories is that this approach does not consider the specific circumstances (i.e., stimuli of uncertainty and applicable moderators) in which the uncertainty is perceived, which could influence whether a specific thought, feeling, or action ultimately helps manage the uncertainty of a particular situation or not. For this reason, the model of uncertainty tolerance for health professions learners (Figure 3.1) makes no such value judgements.

Returning to the example of uncertainty extending from Zach’s chest pain, how might you respond if you were Zach’s nurse? Case Study 3.1b describes some of the cognitive, emotional, and behavioural responses that might occur.

The case includes several responses to uncertainty. Although you as the nurse in this scenario felt worried, this seems concordant with the stakes and acuity of Zach’s presentation. A degree of stress can improve performance, so it may be considered adaptive to managing the uncertainty in this situation (Ilgen et al., 2021). Other adaptive responses in this particular scenario include acknowledging the presence of uncertainty while keeping Zach informed about his management and seeking information through relevant tests. At this stage in the case, it is unclear what the cause of Zach’s chest pain is. A troponin level within the normal range will effectively rule out an acute myocardial infarction, however, the cause of Zach’s chest pain is still unknown.

There are many other ways in which the nurse caring for Zach might have responded to the uncertainty in his case. Instead of prioritising investigations based on clinical assessment, they could have engaged a scattergun approach, whereby investigations that were less relevant or not immediately indicated were arranged, in the hope of lessening the uncertainty. Furthermore, rather than working with Zach to explore the cause of his pain and where uncertainty exists, they could have taken a more paternalistic approach by withholding, rather than disclosing, their uncertainty to Zach. And instead of continuing to care for Zach within their scope of practice, they could have referred Zach to another healthcare professional.

Ultimately, whether a response results in effective, or adaptive, management of uncertainty or not will depend on the specific uncertainty stimulus and the moderators involved. Although many presentations of chest pain result in a diagnosis, it is possible that the nurse in this scenario may not ever find out conclusively what the cause of Zach’s chest pain is, but effective clinical care is not always about diagnostic certainty. Rather, uncertainty tolerance, and indeed high-quality care in general, is about identifying the next best step, adapting and adjusting as more information comes in, and ensuring that the priorities, values, and way of life of the patient are considered and integrated within care plans.

Hillen et al.’s (2017) integrative model of uncertainty tolerance suggests that cognitive, emotional, and behavioural responses equally contribute to determining an individual’s uncertainty tolerance. Research in the context of medical students found that emotional responses to uncertainty are predominantly characterised by a degree of stress (Stephens et al., 2021). This finding is consistent with research by Anderson et al. (2019) in psychology, which suggested that humans may tend to respond to uncertainty, at least initially, with seemingly ‘negative’ emotions, such as anxiety and fear. Further research with medical students has shown that despite these negative emotions, many learners are still able to respond to uncertainty adaptively in the domains of cognition and behaviour (Stephens et al., 2021; Stephens et al. 2023). Hence key definitional aspects of uncertainty ‘tolerance’ maybe one’s ability to recognise and appraise and emotions felt in response to uncertainty, and proceed to respond adaptively by thinking and acting in a manner aligned with evidence-based and person-centred approaches to care (Ilgen et al. 2021, Stephens et al. 2023).

Responses to Uncertainty Over Time

Hillen et al.’s (2017) integrative model of uncertainty tolerance remains a helpful starting point for breaking apart a complex construct, but the model has some limitations. To acknowledge the unclear boundaries of the construct, based on differing conceptual assumptions, the developers of the model include ‘fuzzy boundaries’ and specifically state that they don’t intend the model to act as a ‘grand unifying theory’ of uncertainty tolerance.

In the chapter authors’ studies, Hillen et al.’s (2017) model was used as a starting point to understand how learners experience and respond to uncertainty (Stephens et al., 2021; Stephens et al., 2022b, 2022a, 2023). These studies encountered challenges related to the model in understanding the longitudinal changes that an individual experiences when developing uncertainty tolerance through health professions education, as Hillen et al.’s (2017) integrative model does not explicitly address the moderating factor of time and/or experience. Indeed, these longitudinal studies with medical students identified that some individuals’ responses to uncertainty, as well as the outcomes of uncertain situations, act as moderators. Much like a feedback loop in physiology, and has been reported for other stimulus-response constructs in psychology, current experiences shape future perceptions and responses to uncertainty. For instance, critical reflection can be an adaptive response to uncertainty but can also serve as a moderating factor for future responses to uncertainty (Stephens et al., 2022b). Reflection seems to allow learners to refocus away from emotional responses to uncertainty, such as worry, towards how uncertainty influences and drives learning. Accordingly, reflection can build confidence and an acceptance of uncertainty in healthcare (Stephens et al., 2022b). Although critical reflection is generally an adaptive response and moderator of educational uncertainty, the exception for this may be in relation to traumatic experiences, after which reflection may amplify adverse effects, especially when proximate to the emotions of the event (van der Kolk, 2000). The model of uncertainty tolerance for health professions learners (Figure 3.1) includes a feedback loop, and encounters with future uncertainty stimuli, as understanding how responses to uncertainty can influence future uncertainty is important when building opportunities for learners to practise managing uncertainty through a program of study.

Other Conceptual Models of Uncertainty Tolerance

While this chapter explores in detail Hillen et al.’s (2017) integrative model of uncertainty tolerance and adds the author’s model of uncertainty tolerance for health professions learners, other models are described in the literature (Gerrity et al., 1990; Lee et al., 2021; Scott et al., 2023). These models, including the model of uncertainty tolerance for health professions learners, typically share foundational elements and address similar concepts to those defined by Hillen et al. (2017). For example, Lee et al. (2021) describe a framework of uncertainty for medical education developed following a scoping review of conceptual models of uncertainty. In the graphic representation of this framework, uncertainty is at the centre and is surrounded by subjective influencers of and responses to uncertainty. The representation includes curved and bidirectional arrows both between these elements and extending to and from the centrally placed uncertainty, to indicate that uncertainty is a complex phenomenon. Hence, Hillen et al.’s (2017), the chapter authors’ and Lee et al.’s (2021) models include sources, or stimuli, of uncertainty, as well as factors, or moderators, that influence perceptions of and responses to uncertainty. Each of these models also highlight the complexity of the construct: For example, Hillen et al. (2017) indicate this by using blurred edges in the graphic representation of the model, chapter authors illustrate this with dashed lines and shadowing, and Lee et al. (2021) through the use of circular and bidirectional arrows.

Summary

Conceptual modelling of uncertainty tolerance carries with it a level of uncertainty, but serves as a framework for explaining this complex phenomenon. The key components of the model of uncertainty tolerance for health professions learners (Figure 3.1) can be invaluable for understanding the construct and guiding curriculum development. Health professions educators should consider how the components of the model relate to educational activities, including how uncertainty is stimulated; how learners may respond in terms of their thoughts, feelings, and actions; and the various factors that may influence learners’ perceptions of and responses to uncertainty. This handbook will guide educators in this pursuit.

---

---

References

Ahmed, O. (2014). Processes underlying generalisation of conditioning in humans. [Doctoral thesis, University of New South Wales]. https://doi.org/10.26190/unsworks/2744

Anderson, E. C., Carleton, R. N., Diefenbach, M., & Han, P. K. (2019). The relationship between uncertainty and affect. Frontiers in Psychology, 10, Article 2504. https://doi.org/10.3389/fpsyg.2019.02504

Budner, S. (1962). Tolerance for ambiguity scale. Journal of Personality, 30(1), 29–50.

Diwekar, U., Amekudzi-Kennedy, A., Bakshi, B., Baumgartner, R., Boumans, R., Burger, P., Cabezas, H., Egler, M., Farley, J., Fath, B., Gleason, T., Huang, Y., Karunanithi, A., Khanna, V., Mangan, A., Mayer, A. L., Mukherjee, R., Mullally, G., Rico-Ramirez, V., … Theis, T. (2021). A perspective on the role of uncertainty in sustainability science and engineering. Resources, Conservation and Recycling, 164, Article 105140. https://doi.org/10.1016/j.resconrec.2020.105140

Frenkel-Brunswik, E. (1949). Intolerance of ambiguity as an emotional and perceptual personality variable. Journal of Personality, 18(1), 108–143. https://doi.org/10.1111/j.1467-6494.1949.tb01236.

Funtowicz, S. (2020). From risk calculations to narratives of danger. Climate Risk Management, 27, Article 100212. https://doi.org/10.1016/j.crm.2020.100212

Funtowicz, S. O., & Ravetz, J.R. (1994). Uncertainty, complexity and post‐normal science. Environmental Toxicology and Chemistry: An International Journal, 13(12), 1881–1885.

Gerrity, M. S., DeVellis, R. F., & Earp, J. A. (1990). Physicians’ reactions to uncertainty in patient care: A new measure and new insights. Medical Care, 28(8), 724–736. https://doi.org/10.1097/00005650-199008000-00005

Gupta, P. P., & Fogarty, T. J. (1993). Governmental auditors and their tolerance for ambiguity: An examination of the effects of a psychological variable. Government Accountants Journal, 42(3), 25.

Han, P. K. J., Klein, W. M. P., & Arora, N. K. (2011). Varieties of uncertainty in health care: A conceptual taxonomy. Medical Decision Making, 31(6), 828–838. https://doi.org/10.1177/0272989×11393976

Hillen, M. A., Gutheil, C. M., Strout, T. D., Smets, E. M. A., & Han, P. K. J. (2017). Tolerance of uncertainty: Conceptual analysis, integrative model, and implications for healthcare. Social Science & Medicine, 180, 62–75. https://doi.org/10.1016/j.socscimed.2017.03.024

Ilgen, J. S., Teunissen, P. W., de Bruin, A. B., Bowen, J. L., & Regehr, G. (2021). Warning bells: How clinicians leverage their discomfort to manage moments of uncertainty. Medical Education, 55(2), 233–241. https://doi.org/10.1111/medu.14304

Ilgen, J. S., Watsjold, B. K., & Regehr, G. (2022). Is uncertainty tolerance an epiphenomenon? Medical Education, 56(12), 1150–1152. https://doi.org/10.1111/medu.1493

Lazarus, M. D., & Funtowicz, S. (2023). Learning together: Facing the challenges of sustainability transitions by engaging uncertainty tolerance and post-normal science. Sustainable Earth Reviews, 6(18).  https://doi.org/10.1186/s42055-023-00066-3

Lazarus, M. D., Gouda-Vossos, A., Ziebell, A., & Brand, G. (2022). Fostering uncertainty tolerance in anatomy education: Lessons learned from how humanities, arts and social science (HASS) educators develop learners’ uncertainty tolerance. Anatomical Sciences Education, 16(1), 128–147. https://doi.org/10.1002/ase.2174

Lee, C., Hall, K., Anakin, M., & Pinnock, R. (2021). Towards a new understanding of uncertainty in medical education. Journal of Evaluation in Clinical Practice, 27(5), 1194–1204. https://doi.org/10.1111/jep.13503

Makkawi, B. A., & Rutledge, R. W. (2000). Evaluating audit risk: The effects of tolerance-for-ambiguity, industry characteristics, and experience. In Advances in Accounting Behavioral Research (pp. 69–89). Emerald. https://doi.org/10.1016/S1474-7979(00)03026-X

Patel, P., Hancock, J., Rogers, M., & Pollard, S. R. (2022). Improving uncertainty tolerance in medical students: A scoping review. Medical Education, 56(12), 1163–1173. https://doi.org/10.1111/medu.14873

Pearce, J. M. (1987). A model for stimulus generalization in Pavlovian conditioning. Psychological Review, 94(1), 61–73. https://doi.org/10.1037/0033-295X.94.1.61

Scott, I. A., Doust, J. A., Keijzers, G. B., & Wallis, K. A. (2023). Coping with uncertainty in clinical practice: A narrative review. Medical Journal of Australia, 218(9), 418–425. https://doi.org/10.5694/mja2.51925

Stephens, G. C. (2022). Exploring uncertainty tolerance in medical students [Doctoral thesis, Monash University]. Bridges. https://doi.org/10.26180/21127612.v1

Stephens, G. C., Rees, C. E., & Lazarus, M. D. (2021). Exploring the impact of education on preclinical medical students’ tolerance of uncertainty: A qualitative longitudinal study. Advances in Health Sciences Education, 26(1), 53–77. https://doi.org/10.1007/s10459-020-09971-0

Stephens, G. C., Sarkar, M., & Lazarus, M. D. (2022a). ‘A whole lot of uncertainty’: A qualitative study exploring clinical medical students’ experiences of uncertainty stimuli. Medical Education, 56(7), 736–746. https://doi.org/10.1111/medu.14743

Stephens, G. C., Sarkar, M., & Lazarus, M. D. (2022b). Medical student experiences of uncertainty tolerance moderators: A longitudinal qualitative study. Frontiers in Medicine, 9, Article 864141. https://doi.org/10.3389/fmed.2022.864141

Stephens, G. C., Sarkar, M., & Lazarus, M. D. (2023). ‘I was uncertain, but I was acting on it’: A longitudinal qualitative study of medical students’ responses to uncertainty. Medical Education. Advance online publication. https://doi.org/10.1111/medu.15269

Strout, T. D., Hillen, M., Gutheil, C., Anderson, E., Hutchinson, R., Ward, H., Kay, H., Mills, G. J., & Han, P. K. J. (2018). Tolerance of uncertainty: A systematic review of health and healthcare-related outcomes. Patient Education and Counseling, 101(9), 1518–1537. https://doi.org/10.1016/j.pec.2018.03.030

van der Kolk, B. (2000). Posttraumatic stress disorder and the nature of trauma. Dialogues in Clinical Neuroscience, 2(1), 7–22. https://doi.org/10.31887/DCNS.2000.2.1/bvdkolk

van Iersel, M. B., Brantjes, E., de Visser, M., Looman, N., Bazelmans, E., & van Asselt, D. (2019). Tolerance of clinical uncertainty by geriatric residents: A qualitative study. European Geriatric Medicine, 10(3), 517–522. https://doi.org/10.1007/s41999-019-00199-9

Williams, A.M. & 2012 Baláž, V. (2012). Migration, risk, and uncertainty: Theoretical perspectives. Population, Space and Place, 18(2): 167-180. https://doi.org/10.1002/psp.663