Elsevier

Neuropsychologia

Volume 51, Issue 1, January 2013, Pages 38-51
Neuropsychologia

Waves of regret: A meg study of emotion and decision-making

https://doi.org/10.1016/j.neuropsychologia.2012.10.015Get rights and content

Abstract

Recent fMRI studies have investigated brain activity involved in the feeling of regret and disappointment by manipulating the feedback participants saw after making a decision to play certain gambles: full-feedback (regret: participant sees the outcomes from both the chosen and unchosen gamble) vs. partial-feedback (disappointment: participant only sees the outcome from chosen gamble). However, regret and disappointment are also characterized by differential agency attribution: personal agency for regret, external agency for disappointment. In this study, we investigate the neural correlates of these two characterizations of regret and disappointment using magnetoencephalography (MEG). To do this, we experimentally induced each emotion by manipulating feedback (chosen gamble vs. unchosen gamble), agency (human vs. computer choice) and outcomes (win vs. loss) in a fully randomized design. At the behavioral level the emotional experience of regret and disappointment were indeed affected by both feedback and agency manipulations. These emotions also differentially affect subsequent choices, with regret leading to riskier behavior. At the neural level both feedback and agency affected the brain responses associated with regret and disappointment, demonstrating differential localization in the brain for each. Notably, feedback regret showed greater brain activity in the right anterior and posterior regions, with agency regret producing greater activity in the left anterior region. These findings extend the evidence for neural activity in processing both regret and disappointment by highlighting for the first time the respective importance of feedback and agency, as well as outlining the temporal dynamics of these emotions.

Highlights

► Neural correlates of feedback and agency mechanisms of regret and disappointment. ► Differential processing of regret and disappointment between 190 and 305 ms. ► Larger brain activity for agency regret over the left anterior region. ► Stronger activity for feedback regret over the right anterior and posterior regions.

Introduction

Anyone who has ever made an important decision, such as whether to enter a profession, have children, buy a house, or move abroad, knows that emotions play an important role in decision-making. Indeed, even simple choices, such as how and where to spend the weekend or which university course to take, evoke emotional reactions which depend not only on the outcome itself, but also on how this outcome is achieved. Most decisions involve uncertainty about the consequences of our choice, and therefore before making a decision we often try to think through all possible outcomes, with one of the most important goals of our daily decision-making to choose options that will avoid negative emotional consequences (e.g., Bell, 1982). In the last several decades, the study of how emotions influence decisions has become widespread across several domains, particularly in psychology (e.g., Kahneman and Tversky, 1982, Mellers et al., 1999, Ritov and Baron, 1990, Zeelenberg et al., 1996) and economics (Bell, 1982, Loomes and Sugden, 1982), but also in other fields such as marketing (Inman et al., 1997, Simonson, 1992), cross-cultural psychology (Gilovich, Wang, Regan & Nishina, 2003) and, more recently, cognitive neuroscience (Camille et al., 2004, Chandrasekhar et al., 2008, Chiu et al., 2008, Chua et al., 2009, Coricelli et al., 2005, Liu et al., 2007, Lohrenz et al., 2007, Nicolle et al., 2011a, Nicolle et al., 2011b, Shiv et al., 2005).

A class of negative emotions that seem particularly aversive are those that create the experienced affective state of “if only I had chosen differently”, more commonly known as regret. Thus, before making a decision, people often attempt to anticipate whether they may feel future regret as a consequence of their choice (Bell, 1982, Loomes and Sugden, 1982). In addition to regret, the emotion of disappointment has also been studied extensively across both behavioral science (e.g., Mellers et al., 1999, Zeelenberg and Pieters, 2006) and neuroscience (e.g., Camille et al., 2004, Chua et al., 2009, Coricelli et al., 2005). Disappointment occurs when the actual outcome of a decision is worse than our expectations, and a better outcome would have been possible with a different state of the world (e.g., Bell, 1985, Loomes and Sugden, 1986). Focusing on specific emotions is useful, as research has shown that different emotions have idiosyncratic behavioral tendencies (e.g., Frijda et al., 1989, Roseman et al., 1994). With regard to decision-making, Zeelenberg and Pieters (1999) demonstrated that regret and disappointment are two important emotions which arise in the context of decisions and their associated outcomes, and that are dissociable. While it is certainly true that other emotions can arise after a negative outcome, such as, for example, guilt or shame, regret and disappointment are the two complex emotions that have been studied in most detail with regard to decision-making, likely because both affective states are highly related to the hedonic value of the decision outcomes (e.g., Zeelenberg, van Dijk & Manstead, 1998). Further, emotions such as guilt and shame are more often related to the transgression of a moral or social norm, whereas regret and disappointment are more involved with the ‘pure’ decision outcome itself. Thus, in this study we focused on regret and disappointment to better understand and differentiate the behavioral and neural dynamics associated with each, in order to gain insight into how these emotions can potentially affect our decision strategies.

Regret and disappointment are similar to a certain extent, in that both can arise as a reaction to an unsatisfactory outcome, both are related to decisions taken, and both can even occur simultaneously. Indeed, it is difficult to experience regret without also experiencing disappointment, as both arise when outcomes deviate from one's expectations (e.g., Zeelenberg & Pieters, 2006) and both stem from counterfactual thinking (e.g., Zeelenberg et al., 1998a). Nevertheless, they differ in several important aspects. Regret and disappointment arise from two different counterfactual thoughts: “behavior-focused counterfactuals” for regret and “situation-focused counterfactuals” for disappointment (e.g., van Dijk, Zeelenberg & van der Pligt, 2003), and these two states emerge from a comparison between “what is” and “what might have been”, for example if we had made a different choice (regret), or if another state of the world had occurred (disappointment) (Zeelenberg, van Dijk, Manstead & van der Pligt ,1998b). An example of the two emotions is well captured by the following: “The child is disappointed when the Tooth Fairy forgets his third lost tooth. The child's parents regret the lapse” (Landman, 1993, p. 47). Importantly, there are differences in terms of the responsibility that determine each state. Whereas regret is typically related to self-agency (the agent is responsible for the suboptimal decision outcome) and internal attribution (e.g., Gilovich & Medvec, 1994), disappointment is related to other-agency (some external factor is responsible for the outcome) and external attribution (e.g., Frijda et al., 1989).

Behavioral studies have typically investigated the differences between regret and disappointment by using standard decision scenarios, for example a choice between different sections of the same college class, and then by using manipulation of responsibility: a choice made by the participant themselves to induce personal responsibility and thus potential regret, and a choice made by a computer to induce external responsibility and thus potential disappointment (e.g., Giorgetta et al., 2012, Ordönez and Connolly, 2000, Zeelenberg et al., 1998, Zeelenberg et al., 2000). Moreover, regret is also said to exist when the decision-maker receives no feedback about the rejected option (e.g., Bar-Hillel and Neter, 1995, Connolly and Zeelenberg, 2002, Zeelenberg et al., 1998, Zeelenberg, 1999). Indeed, according to the “Decision Justification Theory”, developed by Connolly and Zeelenberg (2002), there are two different type of regret: comparative (outcome-based) and/or causal (responsibility-based). One (comparative) is associated with the comparison between the outcome deriving from the chosen option and some standard, such as the outcome of the rejected option. The other one (causal) is associated with the feeling of self-blame, responsibility, subjective evaluation of the quality of the decision made, without knowing the outcome of the rejected option.

Neuroscientific studies (e.g., Camille et al., 2004, Chua et al., 2009, Coricelli et al., 2005) have used an alternate method to agency to discriminate between regret and disappointment, namely manipulation of the decision feedback. In these fMRI studies, participants made repeated choices between two gambles. Disappointment was induced by showing participants only the negative outcome of the chosen gamble (partial feedback trials), while regret was created by showing both a bad outcome on the chosen gamble, as well as a good outcome on the unchosen gamble (full feedback trials). These studies showed that the feeling of regret is associated with dorsal anterior cingulate cortex, medial OFC and anterior hippocampus, while disappointment activates middle temporal gyrus and dorsal brainstem (Coricelli et al., 2005). They also showed that both emotional states activate anterior insula, part of dorsomedial prefrontal cortex (BA8 region) and lateral orbitofrontal cortex (Chua et al., 2009), with this activity being stronger for regret. Though neuroscience studies have largely ignored investigating the role of agency on regret and disappointment, Coricelli et al. (2005) and, more recently, Nicolle et al. (2011a) did employ trials where a computer made the choice, but these were only used as a control condition. Specifically, Nicolle et al. (2011a), in a study aimed at investigating how regret affects choice repetition, did not find any behavioral or neural effects associated with computer choices, but did with human choices. Importantly, the crucial role of self-blame and responsibility in the experience of regret, and its role on the subsequent choices has also been recently shown (Nicolle et al., 2011b). However, in this study computer choice and feedback manipulation were not taken into consideration.

Overall therefore, findings from both behavioral science and neuroscience suggest that regret exerts a more substantial influence on choice than disappointment, and also that the emotional impact of regret is stronger than that of disappointment (e.g., Chua et al., 2009, Mellers et al., 1999). However, as both behavioral and neuroscientific approaches have used different means of inducing these emotions (agency-based and feedback-based respectively), an important contribution to the investigation of emotions and how they impact decision-making is to use neuroimaging methods to assess whether both agency and feedback explanations for regret and disappointment emanate from the same underlying psychological and neural process.

To determine the brain processes specifically related to the emotions that play a pivotal role in decision-making, we used magnetoencephalography (MEG). Use of MEG allowed us to explore the temporal dynamics of their experience. Emotions have early and short-lasting effects on the brain (e.g., D'Hondt et al., 2010, Hung et al., 2010, Peyk et al., 2008), and so the superior temporal resolution of MEG enables more precise quantification. Specifically, we employed a well established gambling task paradigm, previously used in fMRI studies on regret and disappointment (e.g., Camille et al., 2004, Chua et al., 2009, Coricelli et al., 2005, Nicolle et al., 2011a), in which we applied both feedback-based and agency-based manipulations. This design therefore allowed us to carefully separate the various components of regret and disappointment. Additionally, previous studies on regret/disappointment (e.g., Camille et al., 2004, Chua et al., 2009, Mellers et al., 1999) have simply asked participants to rate their emotions on a univariate happy/sad scale. This scale provides relatively weak discriminability between regret/disappointment and a general sad mood, and also does not ensure that regret and disappointment are in fact the two most important components of the participants’ experience. To address these points we also asked, during a training phase, about the same happy/sad ratings but also employed a debriefing session where we specifically assessed regret and disappointment in addition to other relevant emotions.

At a behavioral level, we hypothesize that regret and disappointment will be the two primary emotions experienced in our task, with greater differential effect for regret when one is personally responsible for the negative outcome, and for disappointment when one is not. This latter point is of considerable interest in terms of the validity of using gambling tasks when studying regret and disappointment. We also explored the risk-taking behavior and reaction times exhibited by participants and examined the differential effects of experienced emotions on the subsequent choices. We hypothesize that regret, but not disappointment, exerts an important role on subsequent choice behavior. Such a finding would be important in that it can show a converging role for both agency and feedback on choices.

At a neural level, we aim to extend previous findings on regret and disappointment in order to demonstrate that both feedback and agency should be taken into account when investigating the processing of these affective states. We addressed this point by looking separately at the brain's global field activity for the role of feedback and agency respectively. We hypothesize that relatively early after the presentation of the outcome there is separation in their processing. Looking separately at the effect of feedback and agency, using the local field activity level, enabled us to answer another compelling question, whether these two different components (feedback and agency) share the same cortical localization or rather if they are implemented in different brain regions. One possibility is that they are coded in different brain regions. fMRI studies suggest a right brain specialization, based on feedback accounts for regret and disappointment. Thus, one potential hypothesis is that when looking at the feedback effects we will find differential activity, after the full and partial feedback, localized in the right regions of the brain. If shown, this finding will be important as it will strongly support the notion of right hemispheric responses to negative emotions. With regard to agency effects, differential brain activity related to the experience of regret and disappointment would support the idea, stemming from behavioral science, that these two emotions differ based on responsibility attribution. Additionally, different cortical localization between agency and feedback would support the idea of regret as a multi-component process. Though both components (comparative-feedback and causal-agency) have a crucial role in determining the experience of regret (Connolly & Zeelenberg, 2002), to date no study has examined whether these experiences co-occur or are to some degree separable. By assessing the similarities and differences of the neural dynamics of feedback and agency on regret and disappointment, we can significantly contribute to integrating disparate strands of research on these emotions, as well as increasing our understanding of how the brain processes emotions, in particular emotions directly related to decision-making.

Section snippets

Participants

Sixteen right-handed participants (8 Males, mean age: 25.5±3.6 years; mean education: 17.03±2.1 years) participated in the study. Three of these sixteen participants were excluded from the analyses for signal artifacts, and so data is reported from a total of 13 participants. In order to exclude participants with abnormal emotional reactivity, in the assessment phase participants completed the Positive and Negative Affective scales (PANAS, Watson, Clark & Tellegen, 1988). The mean ratings for

Training phase

Fig. 2 (a and b) illustrates results from “Outcome” and “Choice” ratings in all of the four emotion conditions. A 2×2×2 mixed ANOVA analysis [Emotion Valence (Positive vs. Negative) X Agency (Human vs. Computer) X Rating Type (Outcome vs. Choice)] revealed a main effect of Emotion Valence, F(1, 12)=5.3, p<0.05. That is, positive emotions (defined by obtained wins and foregone losses) received stronger emotional ratings than negative emotions (defined by obtained losses and foregone wins). A

Discussion

In this study we investigated the neuronal response to emotions following risky decision-making, primarily in regard to the evoked states of regret and disappointment. In the field of decision-making these are two well-investigated emotions (e.g., Bell, 1982, Bell, 1985, Loomes and Sugden, 1982, Loomes and Sugden, 1986, Mellers et al., 1997, Zeelenberg and Pieters, 2005) which have been shown to play a crucial role in making choices (e.g., Bell, 1982, Bell, 1985, Loomes and Sugden, 1982, Loomes

Conclusion

The present study shows that there is differential electromagnetic activity in the brain associated with the processing of regret and disappointment not only as a function of feedback, as has previously been shown, but also as a function of agency. Tying together the divergent literature on regret and disappointment, these results argue that both the role of feedback (Chua et al., 2009, Mellers et al., 1999), and the role of agency (Ordönez and Connolly, 2000, Zeelenberg et al., 1998,

Acknowledgments

This research was funded by an award to AGS from Provincia Autonoma di Trento. We are grateful for assistance of Filippo Rossi in data analysis.

References (97)

  • Y. Hung et al.

    Unattended emotional faces elicit early lateralized amygdale-frontal and fusiform activations

    NeuroImage

    (2010)
  • M.L. Kringelbach et al.

    The functional neuroanatomy of the human orbitofrontal cortex: evidence from neuroimaging and neuropsychology

    Progress in Neurobiology

    (2004)
  • M. Larquet et al.

    Impaired decision making in schizophrenia and orbitofrontal cortex lesion patients

    Schizophrenia Research

    (2010)
  • J. Marco-Pallarés et al.

    Human oscillatory activity associated to reward processing in a gambling task

    Neuropsychologia

    (2008)
  • L.D. Ordönez et al.

    Regret and responsibility: A reply to Zeelenberg et al., 1998

    Organizational Behavior and Human Decision Processes

    (2000)
  • D. Polezzi et al.

    Brain correlates of risky decision-making

    Neuroimage

    (2010)
  • N.J. Roese et al.

    Schizophrenia involves impairment in the activation of intentions by counterfactual thinking

    Schizophr Res

    (2008)
  • E.K. Silberman et al.

    Hemispheric lateralization of functions related to emotion

    Brain and Cognition

    (1986)
  • W.W. van Dijk et al.

    Blessed are those expected nothing: Lowering expectations as a way of avoiding disappointment

    Journal of Economic Psychology

    (2003)
  • Y. Wu et al.

    The P300 and reward valence, magnitude, and expectancy in outcome evaluation

    Brain Research

    (2009)
  • M. Zeelenberg et al.

    Consequences of regret aversion: Effects of expected feedback on risky decision making

    Organizational Behavior and Human Decision Processes

    (1996)
  • M. Zeelenberg et al.

    Reconsidering the relation between regret and responsibility

    Organizational Behavior and Human Decision Processes

    (1998)
  • M. Zeelenberg et al.

    Emotional reactions to the outcomes of decisions: The role of counterfactual thought in the experience of regret and disappointment

    Organizational Behavior and Human Decision Processes

    (1998)
  • M. Zeelenberg et al.

    Regret and responsibility resolved? Evaluating the conclusions of Ordonez and Connolly 2000

    Organizational Behavior and Human Decision Processes

    (2000)
  • C.J. Anderson

    The psychology of doing nothing: Forms of decision avoidance result from reason and emotion

    Psychological Bulletin

    (2003)
  • M. Bar-Hillel et al.

    Why are people reluctant to exchange lottery tickets?

    Journal of Personality and Social Psychology

    (1995)
  • D.E. Bell

    Regret in decision making under uncertainty

    Operations Research

    (1982)
  • D.E. Bell

    Disappointment in decision making under uncertainty

    Operations Research

    (1985)
  • N. Camille et al.

    The involvement of the orbitofrontal cortex in the experience of regret

    Science

    (2004)
  • N. Canessa et al.

    Understanding others’ regret: A fMRI study

    PLoS One

    (2009)
  • P.H. Chiu et al.

    Smokers’ brains compute, but ignore, a fictive error signal in a sequential investment task

    Nature Neuroscience

    (2008)
  • J.A. Clark et al.

    Living with treatment decisions: regrets and quality of life among men treated for metastatic prostate cancer

    Journal of Clinical Oncology

    (2001)
  • J.A. Coan et al.

    Frontal EEG asymmetry and the behavioral activation and inhibition systems

    Psychophysiology

    (2003)
  • J.A. Coan et al.

    Frontal EEG asymmetry as a moderator and mediator of emotion

    Biological Psychology

    (2004)
  • T. Connolly et al.

    Regret in cancer-related decisions

    Health Psychology

    (2005)
  • T. Connolly et al.

    Regret in decision making

    Current Directions in Psychological Science

    (2002)
  • G. Coricelli et al.

    Regret and its avoidance: A neuroimaging study of choice behavior

    Nature Neuroscience

    (2005)
  • F. D'Hondt et al.

    Early brain-body impact of emotional arousal

    Frontiers in Human Neuroscience

    (2010)
  • R.J. Davidson

    Affect, cognition and hemispheric specialization

  • R.J. Davidson

    Affective style and affective disorders: Perspectives from affective neuroscience

    Cognition and Emotion

    (1998)
  • R.J. Davidson et al.

    Human electroencephalography

  • R.J. Davidson

    What does the prefrontal cortex “do” in affect: Perspectives on frontal EEG asymmetry research

    Biological Psychology

    (2004)
  • F. De Vico Fallani et al.

    Cortical network analysis in patients affected by schizophrenia

    Brain Topography

    (2010)
  • N. Doñamayor et al.

    Temporal dynamics of reward processing revealed by magnetoencephalography

    Human Brain Mapping

    (2011)
  • N. Franck et al.

    Defective recognition of one's own actions in patients with schizophrenia

    American Journal of Psychiatry

    (2001)
  • N.H. Frijda et al.

    Relations among emotion, appraisal, and emotional action readiness

    Journal of Personality and Social Psychology

    (1989)
  • W.J. Gehring et al.

    The medial frontal cortex and the rapid processing of monetary gains and losses

    Science

    (2002)
  • D.T. Gilbert et al.

    Looking forward to looking backward. the misprediction of regret

    Psychological Science

    (2004)
  • Cited by (32)

    • Identifying appropriate service recovery strategies in the event of a natural disaster

      2021, Journal of Hospitality and Tourism Management
      Citation Excerpt :

      Anger has been shown to be a precursor of complaining, negative word of mouth and switching behaviour (Bonifield & Cole, 2007; Mattila & Ro, 2008). Conversely, disappointment, relates to external attribution for the service failure and is typically a reaction to an outcome that is different from what the customer expected (Sanchez - Garcia & Curras - Perez, 2011; Giorgetta et al., 2013). While Zeelenberg et al. (1998) suggest that disappointment may promote negative word of mouth, Zeelenberg and Pieters (2004), concluded that disappointment is less harmful than anger.

    • fMRI in economics: What functional imaging of the brain can add to behavioral economics experiments

      2019, Biophysical Measurement in Experimental Social Science Research: Theory and Practice
    • What are they good for? A constructionist account of counterfactuals in ordinary Chinese

      2017, Journal of Pragmatics
      Citation Excerpt :

      The ubiquity of counterfactual emotions in human experience implies their status as experiential universals. Indeed, their universality is underscored by their neural underpinnings revealed in cognitive neuropsychology and affective neuroscience (Camille et al., 2004; Coricelli et al., 2005; Chua et al., 2009; Liu et al., 2007; Lohrenz et al., 2007; Marchiori and Warglien, 2008; Coricelli and Rustichini, 2010; Nicolle et al., 2011a,b; Giorgetta et al., 2012; Van Hoeck et al., 2013). Given the universality of counterfactual reasoning, it would seem reasonable to assume that as an experiential universal it should be crystallized as a semantic universal in the sense of a basic concept “that must be expressed if language is to be a successful means of communication” (Sapir, 1949:93).2

    • Understanding online regret experience in Facebook use - Effects of brand participation, accessibility &amp; problematic use

      2016, Computers in Human Behavior
      Citation Excerpt :

      The phenomenon of regret experience is well researched in offline environments (Knapp, Stafford, & Daly, 1986; Meyer, 2011; Meyer & Rothenberg, 2004). The prior research examining offline regret experiences focused on three aspects: First, the development of regret experience as a concept (Giorgetta et al., 2013; Inman, 2007; McConnell et al., 2000; Su, Chen, & Zhao, 2008; Zeelenberg & Pieters, 2007; Zeelenberg, van Dijk, van der Pligt, Manstead, van Empelen, & Reinderman, 1998; Zeelenberg et al., 1998; Zeelenberg, 1999); Second, understanding its relationship with consumer decision-making (Bui et al., 2011; Inman et al., 1997; Kang et al., 2009; Lemon et al., 2002; Taylor, 1997; Tsiros & Mittal, 2000); and third, understanding the role of cognitive and behavioral aspects of the offline regret experiences (Giorgetta et al., 2013; Inman, 2007; McConnell et al., 2000; Su et al., 2008; Zeelenberg, van Dijk, & Manstead, 1998; Pieters & Zeelenberg, 2007; Zeelenberg & Pieters, 2007; Zeelenberg et al., 1998; Zeelenberg, 1999). However, only recently have researchers started examining the regret experience in the online environment (Kang et al., 2009; Kaur, Dhir, Chen, & Rajala, 2015; Wang et al., 2011).

    • Understanding online regret experience using the theoretical lens of flow experience

      2016, Computers in Human Behavior
      Citation Excerpt :

      Similarly, 40% of young college students has experienced online regret after sharing online self-portrayal (Stern, 2015). The majority of the prior research concerning regret experiences has addressed consumer decision making (Taylor, 1997; Tsiros & Mittal, 2000; Bui et al., 2011; Kang et al., 2009; Lemon et al., 2002; Inman et al., 1997) and understanding the role of counterfactual thinking, feedback, and cognitive and behavioral aspects of regret regulation for establishing the concept of regret in real life (Zeelenberg, van Dijk, & Manstead, 1998; Pieters & Zeelenberg, 2007; Zeelenberg & Pieters, 2007; McConnell et al., 2000; Su, Chen, & Zhao, 2008; Zeelenberg, van Dijk, van der Pligt, Manstead, van Empelen, & Reinderman, 1998; Zeelenberg, 1999; Inman, 2007; Giorgetta et al., 2013). In comparison, computer-mediated communication researchers have recently begun to examine the reasons and consequences of experiencing online regret.

    • Depressive states amplify both upward and downward counterfactual thinking

      2015, International Journal of Psychophysiology
      Citation Excerpt :

      In this task, participants chose between two options and received positive or negative outcomes of the chosen option and the alternative option (Gu et al., 2011b; Yeung and Sanfey, 2004). Thus, the comparison between positive alternative outcome (AO for short) and negative chosen outcome (CO for short) elicits upward counterfactual thinking, while the comparison between negative AO and positive CO elicits downward counterfactual thinking (Giorgetta et al., 2012). For instance, when a participant discovers that the option he/she selected led to monetary loss (CO) while the other option could have resulted in monetary gain (AO), he/she may feel regret and imagine the preferable alternate selection (upward counterfactual thinking).

    View all citing articles on Scopus
    View full text