Kurzban et al.'s opportunity cost model of mental effort relies heavily on counterfactual thinking – that is, consideration of opportunities missed while the individual performs a task. Missed opportunities are commonly associated with aversive feelings of regret, induced by the knowledge that something might have been better had we chosen to act differently. Furthermore, when multiple possible alternatives are available to us and when we have control over our actions, the anticipation of possible future regret has a strong influence on our choice behaviour (Mellers et al. Reference Mellers, Schwartz and Ritov1999). A regret-averse individual will select behaviours with the aim of minimizing future regret (Savage Reference Savage1951). Such decisions include whether to persist with a current task, or to devote a portion of our computational resources towards alternative mental activities (such as daydreaming or future planning), or even to abandon the task altogether.
The magnitude of anticipated regret is commonly quantified as the difference between the received outcomes of our actual choice and the best possible outcome that might have been received from a different choice (Bell Reference Bell1982; Loomes & Sugden Reference Loomes and Sugden1982). This formalisation is also supported by the neuroimaging evidence that, in brain regions commonly associated with reward and decision-making, activity changes with the magnitude of this measure of regret (Coricelli et al. Reference Coricelli, Critchley, Joffily, O'Doherty, Sirigu and Dolan2005; Nicolle et al. Reference Nicolle, Bach, Frith and Dolan2011a). Accordingly, the anticipated regret associated with performing a given a task can be computed as the expected reward of doing well in the task, minus the best possible reward that might be achieved from devoting some (or all) of our efforts towards an alternative task.
We suggest that Kurzban et al.'s model may be complicated by the commonly observed finding that regret is stronger when arising from decisions to act (or switch task) than from decisions to refrain from acting (or stick with a current task) (e.g., Kahneman & Tversky Reference Kahneman, Tversky, Kahneman, Slovic and Tversky1982). Proposed explanations for this action/inaction asymmetry in regret include suggestions that active decisions are perceived as more directly causal of their consequences, and that passive decisions are perceived as more easily justified than are active decisions. A learned action/inaction asymmetry in the anticipation of regret is also thought to promote a bias towards more passive decisions, such as decision delay, inaction, or sticking with the norm or status quo (e.g., Baron & Ritov Reference Baron and Ritov1994). In further support of this regret-induced status quo bias, neuroimaging data show that activity in the anterior insula (a region commonly involved in error processing and its impact on behaviour) is stronger for action regrets than for inaction regrets, and that this activity is associated with enhanced subsequent inaction bias (Nicolle et al. Reference Nicolle, Fleming, Bach, Driver and Dolan2011b).
On the surface, this action/inaction asymmetry in regret appears to present a problem for Kurzban et al.'s opportunity cost model of mental effort. Their model presents the attractiveness of performing alternative tasks as critical to feelings of mental effort and as a key motivator for withdrawing processing capacity away from the current task. In stark contrast, action/inaction asymmetries in regret predict that anticipated regret is greater for switching away from the current task than for sticking with it, and that this would motivate persistence in the task (if the individual anticipates any regret associated with switching). If we consider a decision to switch computational resources towards an alternative task as an active process, and the decision to persist with the task at hand as a more passive process, then the former should, according to the above findings, be associated with higher anticipated regret. In other words, whereas missed opportunities may present a cost to the individual while performing a given task, decisions to give up and switch to an alternative task may result in greater, regret-based costs. A regret-induced status quo bias would then predict a reduced tendency to switch task.
Although these two costs (i.e., missed opportunities from alternative tasks and the anticipated regret of switching) might be predicted to motivate opposite behavioural tendencies, it seems likely that both types of costs enter into calculations of the benefits of persisting with a given task. The relative influence of each of these costs on our behaviour, however, may depend upon several factors. For example, the possibility for the individual to receive performance-related rewards in the current task may increase the anticipated regret associated with quitting the task, while opportunity costs (associated with the value of alternative tasks) would be unaffected. Therefore, if the individual is offered financial incentive for successfully completing the task, the anticipated regret of missing that reward would be a strong motivator, even if there is a possibility for greater reward from switching to an alternative task. Another important factor may be the individual's uncertainty about the likelihood of reward from the alternative tasks. If uncertainty is high, the anticipated regret of switching task may be enhanced compared to the relative safety of continuing. Finally, the role of anticipated regret in the decision likely depends upon the responsibility the individual has for the consequences of their actions (since personal responsibility is critical for the experience of regret).
In sum, we suggest that some consideration of regret (and particularly of action/inaction asymmetries in anticipated regret) may be necessary for development of an opportunity cost model of subjective mental effort.
Kurzban et al.'s opportunity cost model of mental effort relies heavily on counterfactual thinking – that is, consideration of opportunities missed while the individual performs a task. Missed opportunities are commonly associated with aversive feelings of regret, induced by the knowledge that something might have been better had we chosen to act differently. Furthermore, when multiple possible alternatives are available to us and when we have control over our actions, the anticipation of possible future regret has a strong influence on our choice behaviour (Mellers et al. Reference Mellers, Schwartz and Ritov1999). A regret-averse individual will select behaviours with the aim of minimizing future regret (Savage Reference Savage1951). Such decisions include whether to persist with a current task, or to devote a portion of our computational resources towards alternative mental activities (such as daydreaming or future planning), or even to abandon the task altogether.
The magnitude of anticipated regret is commonly quantified as the difference between the received outcomes of our actual choice and the best possible outcome that might have been received from a different choice (Bell Reference Bell1982; Loomes & Sugden Reference Loomes and Sugden1982). This formalisation is also supported by the neuroimaging evidence that, in brain regions commonly associated with reward and decision-making, activity changes with the magnitude of this measure of regret (Coricelli et al. Reference Coricelli, Critchley, Joffily, O'Doherty, Sirigu and Dolan2005; Nicolle et al. Reference Nicolle, Bach, Frith and Dolan2011a). Accordingly, the anticipated regret associated with performing a given a task can be computed as the expected reward of doing well in the task, minus the best possible reward that might be achieved from devoting some (or all) of our efforts towards an alternative task.
We suggest that Kurzban et al.'s model may be complicated by the commonly observed finding that regret is stronger when arising from decisions to act (or switch task) than from decisions to refrain from acting (or stick with a current task) (e.g., Kahneman & Tversky Reference Kahneman, Tversky, Kahneman, Slovic and Tversky1982). Proposed explanations for this action/inaction asymmetry in regret include suggestions that active decisions are perceived as more directly causal of their consequences, and that passive decisions are perceived as more easily justified than are active decisions. A learned action/inaction asymmetry in the anticipation of regret is also thought to promote a bias towards more passive decisions, such as decision delay, inaction, or sticking with the norm or status quo (e.g., Baron & Ritov Reference Baron and Ritov1994). In further support of this regret-induced status quo bias, neuroimaging data show that activity in the anterior insula (a region commonly involved in error processing and its impact on behaviour) is stronger for action regrets than for inaction regrets, and that this activity is associated with enhanced subsequent inaction bias (Nicolle et al. Reference Nicolle, Fleming, Bach, Driver and Dolan2011b).
On the surface, this action/inaction asymmetry in regret appears to present a problem for Kurzban et al.'s opportunity cost model of mental effort. Their model presents the attractiveness of performing alternative tasks as critical to feelings of mental effort and as a key motivator for withdrawing processing capacity away from the current task. In stark contrast, action/inaction asymmetries in regret predict that anticipated regret is greater for switching away from the current task than for sticking with it, and that this would motivate persistence in the task (if the individual anticipates any regret associated with switching). If we consider a decision to switch computational resources towards an alternative task as an active process, and the decision to persist with the task at hand as a more passive process, then the former should, according to the above findings, be associated with higher anticipated regret. In other words, whereas missed opportunities may present a cost to the individual while performing a given task, decisions to give up and switch to an alternative task may result in greater, regret-based costs. A regret-induced status quo bias would then predict a reduced tendency to switch task.
Although these two costs (i.e., missed opportunities from alternative tasks and the anticipated regret of switching) might be predicted to motivate opposite behavioural tendencies, it seems likely that both types of costs enter into calculations of the benefits of persisting with a given task. The relative influence of each of these costs on our behaviour, however, may depend upon several factors. For example, the possibility for the individual to receive performance-related rewards in the current task may increase the anticipated regret associated with quitting the task, while opportunity costs (associated with the value of alternative tasks) would be unaffected. Therefore, if the individual is offered financial incentive for successfully completing the task, the anticipated regret of missing that reward would be a strong motivator, even if there is a possibility for greater reward from switching to an alternative task. Another important factor may be the individual's uncertainty about the likelihood of reward from the alternative tasks. If uncertainty is high, the anticipated regret of switching task may be enhanced compared to the relative safety of continuing. Finally, the role of anticipated regret in the decision likely depends upon the responsibility the individual has for the consequences of their actions (since personal responsibility is critical for the experience of regret).
In sum, we suggest that some consideration of regret (and particularly of action/inaction asymmetries in anticipated regret) may be necessary for development of an opportunity cost model of subjective mental effort.