commentaries
‘But how do you really feel?’ Measuring emotions in medical education research Anthony R Artino Jr.1 & Laura M Naismith2 ‘From an intuitive layperson perspective, it should be easy to determine whether someone is experiencing a particular emotion. However, scientific evidence suggests that measuring a person’s emotional state is one of the most vexing problems in affective science.’ Mauss & Robinson1 Humans are emotional beings and, although the computer may have served as a useful model for understanding how the human mind processes information, the last two decades have seen the emergence of a new view of human thinking and learning as being much more fuzzy, flexible and subject to the effects of emotion. Today, findings from multiple disciplines support the idea that emotion is intimately and inseparably intertwined with cognition in guiding learning, behaviour, decision making, and more.2–
tion. However, whether we as researchers explore emotions as physiological processes or as social practices,6 we face the common dilemma of how to operationalise this multifaceted construct. In this commentary, we define emotion and describe several contemporary methods for measuring it. Our take-home message is that an understanding of emotion and its role in human functioning – be it identity development or clinical skills development – is closely tied to the way in which we define and measure emotion in our research. By carefully considering the benefits and limitations of various measurement approaches, we can help to ensure that our findings illuminate rather than obscure.
medical student might experience. Our definition differentiates an emotion from a mood, which is usually longer, more diffuse (and thus less intense), and without particular referent. In the education and psychology literatures, moods and emotions are typically considered affective states, as opposed to affective traits (such as neuroticism).1,8 Furthermore, although not a focus of our own work, we also accept that social and cultural norms exert strong pressures on how emotions can be expressed or even acknowledged in various social contexts.6
The last two decades have seen the emergence of a new view of human thinking as much more fuzzy, flexible and subject to the effects of emotion
For most of the 20th century, emotions and their potential impacts on human functioning were largely ignored by education scholars. For instance, psychologists who studied human learning and performance tended to favour cognition over affect and developed theories in which thinking and learning were viewed as information processing, whereas emotion was ignored or otherwise marginalised.2,9 In the last two decades, however, emotions have played a more prominent role in the study of human learning and performance. Today, most theories of learning recognise the inextricably close links between emotion and cognition. In medical education, many scholars now acknowledge, in very explicit ways, that learning to become a doctor is an emotionally charged activity.5,6,8,10
4
There are many ways to study emotion. In this issue of Medical Education, Dornan et al.5 adopt a thoughtful and much-needed sociocultural perspective to explore the links between emotion and identity development in clerkship educa1
Bethesda, Maryland, USA Toronto, Ontario, Canada
2
Correspondence: Anthony R Artino, Jr., Department of Medicine, Uniformed Services University of the Health Sciences, 4301 Jones Bridge Road, Bethesda, Maryland 20814, USA. Tel: 00 1 301 295 3693; E-mails: [email protected] doi: 10.1111/medu.12642
140
The study of any psychological construct requires careful consideration of its definition. Unfortunately, the study of emotion has a long history of definitional disorder. The confusion stems, in part, from the fact that emotions have been studied from many different scholarly perspectives and research traditions. In our own work, and for the purpose of this commentary, we define emotion as a psycho-physiological change that is short-lived, intense and specific to a personally meaningful stimulus.7 For example, anxiety about a looming clinical encounter is an emotion a typical
The study of any psychological construct requires careful consideration of its definition
ª 2015 John Wiley & Sons Ltd. MEDICAL EDUCATION 2015; 49: 138–146
commentaries
In the last two decades, emotions have played a more prominent role in the study of human learning and performance
Hence, if we want to rigorously study the science of emotion, an important question is how we measure it. Dornan et al.5 chose to measure emotion by analysing what students said. Such self-report approaches are appealing and popular because they do not require any special equipment to collect the data, and they provide an opportunity to gain insight into both what participants experience and how they interpret those experiences. However, self-report methods are subject to at least two significant limitations: people differ greatly in their awareness of their own emotional states, and, even with sufficient awareness, they may be unwilling to report on certain types of emotional experiences.1 For example, in their study of identity development, Dornan et al.5 noted that they collected relatively few examples of negative emotional experiences. Was this because the participants: (i) had more positive than negative experiences; (ii) did not have the same insight into their negative experiences as they did into their positive experiences, or (iii) were reluctant to report negative emotions for reasons of self-preservation? They (and we) can only speculate. People differ greatly in their awareness of their own emotional states, and, even with sufficient awareness, may be unwilling to report on certain types of emotional experience
As emotions trigger physiological reactions within the body that can be externally observed and assessed, another measurement
approach focuses on physiology. Skin conductance and heart rate are two examples of indices related to autonomic nervous system activation. While not exclusively triggered as a function of emotional response, changes in physiological indices have shown relationships to emotional dimensions such as valence (positive or negative emotion) and arousal (activating or deactivating emotion).1 Levels of these emotional dimensions can then be used independently to predict behavioural responses. For example, level of arousal has been associated with learning-related behaviours such as receptivity to feedback.10,11 As emotions trigger physiological reactions within the body that can be externally observed, another approach focuses on physiology
Behaviour itself can also be a measure of emotion. We signal emotions to each other through our voices, our facial expressions and even our posture. A considerable body of research over a number of decades has identified specific features and behavioural characteristics of different emotions across cultures.1 This work is now being applied in the study of how emotional processes and learning unfold and interact in real time. For example, Duffy et al.12 are currently using a combination of body language, paralinguistics (vocal pitch, loudness, rate, fluency), dialogue and facial expressions to study emotions and learning during simulation training in emergency medicine. Because emotions are short-lived and specific to personally meaningful stimuli, they are, by most definitions, dynamic and everchanging. As such, it could be argued that a more complete picture of emotion and its effects on
ª 2015 John Wiley & Sons Ltd. MEDICAL EDUCATION 2015; 49: 138–146
human functioning should include contextual, real-time measures. We signal emotions to each other through our voices, our facial expressions and even our posture
Finally, as Dornan et al.5 acknowledge, ethical considerations are of paramount importance when studying emotions. In some cases, measuring participants’ emotions involves more than minimal risk. Thus, informed consent processes should seek to address participants’ understanding of which methods will be used and the implications these have for the kinds of interpretations that will be made. For instance, Dornan et al.5 may have been able to gather further insights by analysing the vocal characteristics of their participants in addition to what they said; however, this may have adversely affected recruitment by potentially increasing the likelihood that participants might be individually identified by the research team. Clearly, such considerations must be weighed carefully. A more complete picture of emotion and its effects on human functioning should include contextual, real-time measures
Our goal in this commentary was to stimulate a thoughtful discussion about the study of emotion in medical education. We suggest that different measures of emotion highlight different facets of the construct. From a practical standpoint, this means that there is no reference standard measure of emotion.1 Therefore, medical education researchers must explicitly define what they mean by emotion and then choose their measurement approaches accordingly.
141
commentaries Ultimately, we believe that careful consideration of the measurement issues identified in this commentary will encourage even more high-quality research into the complex interplay between emotion and a range of diverse psychosocial constructs in medical education.
REFERENCES 1 Mauss IB, Robinson MD. Measures of emotion: a review. Cogn Emot 2009;23 (2):209–37. 2 Picard RW, Papert S, Bender W, Blumberg B, Breazeal C, Cavallo D, Machover T, Resnick M, Roy D, Strohecker C. Affective learning – a manifesto. BT Tech J 2004;22 (4):253–69. 3 Schutz PA, Pekrun R, eds. Emotion in Education. San Diego, CA: Academic Press 2007.
4 Franks DD. The neuroscience of emotions. In: Stets J, Turner JH, eds. Handbook of the Sociology of Emotions. New York, NY: Springer 2006;38–62. 5 Dornan T, Pearson E, Carson P, Helmich E, Bundy C. Emotions and identity in the figured world of becoming a doctor. Med Educ 2015;49:174–85. 6 McNaughton N. Discourse(s) of emotion within medical education: the ever-present absence. Med Educ 2013;47:71–9. 7 Artino Jnr AR. When I say . . . emotion in medical education. Med Educ 2013;47:1062–3. 8 Artino AR, Holmboe E, Durning SJ. Can achievement emotions be used to better understand motivation, learning, and performance in medical education? Med Teach 2012;34 (3):240–4. 9 Dai DY, Sternberg RJ, eds. Motivation, Emotion and Cognition:
Integrative Perspectives on Intellectual Functioning and Development. Mahwah, NJ: Lawrence Erlbaum Associates 2004. 10 McConnell MM, Eva KW. The role of emotion in learning and transfer of clinical skill and knowledge. Acad Med 2012;87 (10):1316–22. 11 Naismith LM. Examining motivational and emotional influences on medical students’ attention to feedback in a technology-rich environment for learning clinical reasoning. PhD dissertation. Montreal, QC: McGill University 2013. 12 Duffy MC, Azevedo R, Griscom S, Lajoie S, Lachapelle K. Measuring emotions in medical simulations: methods, applications, and challenges of a process data approach. Abstract presented at the Royal College Simulation Summit, 12 September 2014, Toronto, ON.
Learning from erroneous examples in medical education Tamara van Gog The article by Domuracki et al.1 published in this issue fits into a long tradition of research on observational learning, also known in educational research as examplebased learning, which is subject to increasing investigation in medical education. Given the aim of improving central line insertion skills, the authors1 understandably focus their discussion of prior research mainly on the observational learning of (psycho)motor
skills. Nevertheless, this study also has interesting parallels with research on training in cognitive tasks through example-based learning. Therefore, I will start by placing this study in the broader context of research on examplebased learning in the educational sciences. Subsequently, I will discuss some factors that may play a role in learning from erroneous examples, which may also provide inspiration for future research on this topic.
Rotterdam, The Netherlands
Correspondence : Tamara van Gog, Institute of Psychology, Erasmus University Rotterdam, PO Box 1738, 3000 DR Rotterdam, The Netherlands. Tel: 00 31 10 408 9041; E-mail: [email protected]
This study has interesting commonalities with research on training in cognitive tasks through example-based learning
Example-based learning has been studied from the perspectives of two different research traditions.2 Research inspired by social learning theory3 has mostly focused on modelling examples, which provide students with opportunities to observe an adult or a peer model performing the learning task, either live (face to face) or on video. Research inspired by ACT-R (adaptive control of thought– rational) theory4 and cognitive load theory5 has mainly focused on worked examples, which typically provide students with a step-by-step written account of how to complete the learning task. Despite these origins in different research traditions, there are many interesting
doi: 10.1111/medu.12655
142
ª 2015 John Wiley & Sons Ltd. MEDICAL EDUCATION 2015; 49: 138–146
Copyright of Medical Education is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
‘But how do you really feel?’ Measuring emotions in medical education research Anthony R Artino Jr.1 & Laura M Naismith2 ‘From an intuitive layperson perspective, it should be easy to determine whether someone is experiencing a particular emotion. However, scientific evidence suggests that measuring a person’s emotional state is one of the most vexing problems in affective science.’ Mauss & Robinson1 Humans are emotional beings and, although the computer may have served as a useful model for understanding how the human mind processes information, the last two decades have seen the emergence of a new view of human thinking and learning as being much more fuzzy, flexible and subject to the effects of emotion. Today, findings from multiple disciplines support the idea that emotion is intimately and inseparably intertwined with cognition in guiding learning, behaviour, decision making, and more.2–
tion. However, whether we as researchers explore emotions as physiological processes or as social practices,6 we face the common dilemma of how to operationalise this multifaceted construct. In this commentary, we define emotion and describe several contemporary methods for measuring it. Our take-home message is that an understanding of emotion and its role in human functioning – be it identity development or clinical skills development – is closely tied to the way in which we define and measure emotion in our research. By carefully considering the benefits and limitations of various measurement approaches, we can help to ensure that our findings illuminate rather than obscure.
medical student might experience. Our definition differentiates an emotion from a mood, which is usually longer, more diffuse (and thus less intense), and without particular referent. In the education and psychology literatures, moods and emotions are typically considered affective states, as opposed to affective traits (such as neuroticism).1,8 Furthermore, although not a focus of our own work, we also accept that social and cultural norms exert strong pressures on how emotions can be expressed or even acknowledged in various social contexts.6
The last two decades have seen the emergence of a new view of human thinking as much more fuzzy, flexible and subject to the effects of emotion
For most of the 20th century, emotions and their potential impacts on human functioning were largely ignored by education scholars. For instance, psychologists who studied human learning and performance tended to favour cognition over affect and developed theories in which thinking and learning were viewed as information processing, whereas emotion was ignored or otherwise marginalised.2,9 In the last two decades, however, emotions have played a more prominent role in the study of human learning and performance. Today, most theories of learning recognise the inextricably close links between emotion and cognition. In medical education, many scholars now acknowledge, in very explicit ways, that learning to become a doctor is an emotionally charged activity.5,6,8,10
4
There are many ways to study emotion. In this issue of Medical Education, Dornan et al.5 adopt a thoughtful and much-needed sociocultural perspective to explore the links between emotion and identity development in clerkship educa1
Bethesda, Maryland, USA Toronto, Ontario, Canada
2
Correspondence: Anthony R Artino, Jr., Department of Medicine, Uniformed Services University of the Health Sciences, 4301 Jones Bridge Road, Bethesda, Maryland 20814, USA. Tel: 00 1 301 295 3693; E-mails: [email protected] doi: 10.1111/medu.12642
140
The study of any psychological construct requires careful consideration of its definition. Unfortunately, the study of emotion has a long history of definitional disorder. The confusion stems, in part, from the fact that emotions have been studied from many different scholarly perspectives and research traditions. In our own work, and for the purpose of this commentary, we define emotion as a psycho-physiological change that is short-lived, intense and specific to a personally meaningful stimulus.7 For example, anxiety about a looming clinical encounter is an emotion a typical
The study of any psychological construct requires careful consideration of its definition
ª 2015 John Wiley & Sons Ltd. MEDICAL EDUCATION 2015; 49: 138–146
commentaries
In the last two decades, emotions have played a more prominent role in the study of human learning and performance
Hence, if we want to rigorously study the science of emotion, an important question is how we measure it. Dornan et al.5 chose to measure emotion by analysing what students said. Such self-report approaches are appealing and popular because they do not require any special equipment to collect the data, and they provide an opportunity to gain insight into both what participants experience and how they interpret those experiences. However, self-report methods are subject to at least two significant limitations: people differ greatly in their awareness of their own emotional states, and, even with sufficient awareness, they may be unwilling to report on certain types of emotional experiences.1 For example, in their study of identity development, Dornan et al.5 noted that they collected relatively few examples of negative emotional experiences. Was this because the participants: (i) had more positive than negative experiences; (ii) did not have the same insight into their negative experiences as they did into their positive experiences, or (iii) were reluctant to report negative emotions for reasons of self-preservation? They (and we) can only speculate. People differ greatly in their awareness of their own emotional states, and, even with sufficient awareness, may be unwilling to report on certain types of emotional experience
As emotions trigger physiological reactions within the body that can be externally observed and assessed, another measurement
approach focuses on physiology. Skin conductance and heart rate are two examples of indices related to autonomic nervous system activation. While not exclusively triggered as a function of emotional response, changes in physiological indices have shown relationships to emotional dimensions such as valence (positive or negative emotion) and arousal (activating or deactivating emotion).1 Levels of these emotional dimensions can then be used independently to predict behavioural responses. For example, level of arousal has been associated with learning-related behaviours such as receptivity to feedback.10,11 As emotions trigger physiological reactions within the body that can be externally observed, another approach focuses on physiology
Behaviour itself can also be a measure of emotion. We signal emotions to each other through our voices, our facial expressions and even our posture. A considerable body of research over a number of decades has identified specific features and behavioural characteristics of different emotions across cultures.1 This work is now being applied in the study of how emotional processes and learning unfold and interact in real time. For example, Duffy et al.12 are currently using a combination of body language, paralinguistics (vocal pitch, loudness, rate, fluency), dialogue and facial expressions to study emotions and learning during simulation training in emergency medicine. Because emotions are short-lived and specific to personally meaningful stimuli, they are, by most definitions, dynamic and everchanging. As such, it could be argued that a more complete picture of emotion and its effects on
ª 2015 John Wiley & Sons Ltd. MEDICAL EDUCATION 2015; 49: 138–146
human functioning should include contextual, real-time measures. We signal emotions to each other through our voices, our facial expressions and even our posture
Finally, as Dornan et al.5 acknowledge, ethical considerations are of paramount importance when studying emotions. In some cases, measuring participants’ emotions involves more than minimal risk. Thus, informed consent processes should seek to address participants’ understanding of which methods will be used and the implications these have for the kinds of interpretations that will be made. For instance, Dornan et al.5 may have been able to gather further insights by analysing the vocal characteristics of their participants in addition to what they said; however, this may have adversely affected recruitment by potentially increasing the likelihood that participants might be individually identified by the research team. Clearly, such considerations must be weighed carefully. A more complete picture of emotion and its effects on human functioning should include contextual, real-time measures
Our goal in this commentary was to stimulate a thoughtful discussion about the study of emotion in medical education. We suggest that different measures of emotion highlight different facets of the construct. From a practical standpoint, this means that there is no reference standard measure of emotion.1 Therefore, medical education researchers must explicitly define what they mean by emotion and then choose their measurement approaches accordingly.
141
commentaries Ultimately, we believe that careful consideration of the measurement issues identified in this commentary will encourage even more high-quality research into the complex interplay between emotion and a range of diverse psychosocial constructs in medical education.
REFERENCES 1 Mauss IB, Robinson MD. Measures of emotion: a review. Cogn Emot 2009;23 (2):209–37. 2 Picard RW, Papert S, Bender W, Blumberg B, Breazeal C, Cavallo D, Machover T, Resnick M, Roy D, Strohecker C. Affective learning – a manifesto. BT Tech J 2004;22 (4):253–69. 3 Schutz PA, Pekrun R, eds. Emotion in Education. San Diego, CA: Academic Press 2007.
4 Franks DD. The neuroscience of emotions. In: Stets J, Turner JH, eds. Handbook of the Sociology of Emotions. New York, NY: Springer 2006;38–62. 5 Dornan T, Pearson E, Carson P, Helmich E, Bundy C. Emotions and identity in the figured world of becoming a doctor. Med Educ 2015;49:174–85. 6 McNaughton N. Discourse(s) of emotion within medical education: the ever-present absence. Med Educ 2013;47:71–9. 7 Artino Jnr AR. When I say . . . emotion in medical education. Med Educ 2013;47:1062–3. 8 Artino AR, Holmboe E, Durning SJ. Can achievement emotions be used to better understand motivation, learning, and performance in medical education? Med Teach 2012;34 (3):240–4. 9 Dai DY, Sternberg RJ, eds. Motivation, Emotion and Cognition:
Integrative Perspectives on Intellectual Functioning and Development. Mahwah, NJ: Lawrence Erlbaum Associates 2004. 10 McConnell MM, Eva KW. The role of emotion in learning and transfer of clinical skill and knowledge. Acad Med 2012;87 (10):1316–22. 11 Naismith LM. Examining motivational and emotional influences on medical students’ attention to feedback in a technology-rich environment for learning clinical reasoning. PhD dissertation. Montreal, QC: McGill University 2013. 12 Duffy MC, Azevedo R, Griscom S, Lajoie S, Lachapelle K. Measuring emotions in medical simulations: methods, applications, and challenges of a process data approach. Abstract presented at the Royal College Simulation Summit, 12 September 2014, Toronto, ON.
Learning from erroneous examples in medical education Tamara van Gog The article by Domuracki et al.1 published in this issue fits into a long tradition of research on observational learning, also known in educational research as examplebased learning, which is subject to increasing investigation in medical education. Given the aim of improving central line insertion skills, the authors1 understandably focus their discussion of prior research mainly on the observational learning of (psycho)motor
skills. Nevertheless, this study also has interesting parallels with research on training in cognitive tasks through example-based learning. Therefore, I will start by placing this study in the broader context of research on examplebased learning in the educational sciences. Subsequently, I will discuss some factors that may play a role in learning from erroneous examples, which may also provide inspiration for future research on this topic.
Rotterdam, The Netherlands
Correspondence : Tamara van Gog, Institute of Psychology, Erasmus University Rotterdam, PO Box 1738, 3000 DR Rotterdam, The Netherlands. Tel: 00 31 10 408 9041; E-mail: [email protected]
This study has interesting commonalities with research on training in cognitive tasks through example-based learning
Example-based learning has been studied from the perspectives of two different research traditions.2 Research inspired by social learning theory3 has mostly focused on modelling examples, which provide students with opportunities to observe an adult or a peer model performing the learning task, either live (face to face) or on video. Research inspired by ACT-R (adaptive control of thought– rational) theory4 and cognitive load theory5 has mainly focused on worked examples, which typically provide students with a step-by-step written account of how to complete the learning task. Despite these origins in different research traditions, there are many interesting
doi: 10.1111/medu.12655
142
ª 2015 John Wiley & Sons Ltd. MEDICAL EDUCATION 2015; 49: 138–146
Copyright of Medical Education is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
