Andrew Westbrook
Intuition in Decision Making
Intuition plays a central role in everyday decision making. Often, when presented with a decision, we have anintuitive sense that one alternative is better than others without being able to report our reasons for feeling so. Under certain circumstances, we make intuitive judgments which can lead to either efficient, adaptive outcomes or misleading, maladaptive outcomes, as in the case of racial profiling.In all of these instances, we may learn to trust or reject our intuitive feelings based on their past performance in leading us to optimal decisions in these various settings.
The defining feature of intuition is that it bridges conscious and non-conscious decision making processes. Though the products of intuition, such as a feeling, urge or hunch, may be available for conscious access, the antecedents of intuition are inaccessible. The degree to which our reasoning is conscious is important for the quality of our decisions because consciousness confers distinct functional advantages such as cognitive flexibility (Price & Norman, 2008), but may also confer comparative disadvantages such as limited cognitive capacity leading to a greater reliance on heuristics (Dijksterhuis, 2006).
Because intuitive reasoning comes to bear on the quality of our decisions, the extent to which we should rely on our intuition has become a central question in the decision making literature. For example, in the "dual process" theory of reasoning, decision theorists study a class of decision outcomes which are arrived at via quick, automatic, and effortless processes, sometimes called “intuitive”, which have been referred to as “System 1” (S1) processes, and another class which requires more deliberate, slow and methodical reasoning, referred to as “System 2” (S2) processes (Stanovich, 1999).
The tradeoff between S1 and S2 is that while subjects using S1 processes may arrive at a solution more rapidly and at a lower cost of computation, S1-type processing has been associated with systematic heuristics and biases which may compromise our reasoning (Tversky & Kahneman, 1974). These biases are key evidence for the widely held view that the best decisions are made using fully conscious reasoning and that semi-conscious intuitive processes are inferior bases for decisions and judgments. However, it is not entirely clear how consciousness, and therefore intuition, maps onto the S1/S2 dichotomy (Price & Norman, 2008). For example,decision tasks intended to study systematic biases require subjects to use a combination of conscious and non-conscious processes and it is not clear that the tasks measure the performance of intuition (Betsch, 2007). Thus, it is wrong to blame biases solely on intuitive reasoning.
When decision tasks are designed to isolate intuitive processes, intuition can lead to comparatively better decision outcomes under certain conditions. In a recent example involving a relatively complex, multi-attribute decision task, subjects relying more on intuitive reasoning made better choices among a set of items to purchase than subjects relying more on conscious deliberation (Dijksterhuis, 2006). The explanation of this finding was that in a complex multi-attribute problem(involving more than one or two attributes and their respective values) subjects who deliberate before coming to a decision are handicapped by the relatively limited capacity of conscious cognition and rely more on the salience of particular attributes, and therefore non-compensatory reasoning strategies. Subjects who rely more on intuition use a more “holistic” evaluative process which results in a more balanced, compensatory weighting of attributes and therefore better decisions.
Converging evidence for this explanation comes from a study in which subjects were tasked with deciding between art posters which they would be allowed to take home for free and were either allowed to take a poster, no questions asked, or were told they were going to be required to explain their choice (Wilson et al., 1993). In this study, subjects required to explain their choice were significantly less satisfied with their choice in the long run. Apparently, asking subjects to generate conscious, explicit reasons for their choice forced them to focus on reasons that were “easy to verbalize”. This availability-type heuristic led to inferior choices in cases of conscious reasoning as compared with semi-conscious intuitive reasoning which seemed to take better account of attributes which were harder to verbalize.
The decision theory answer to the problem of non-compensatory reasoning is Multi-Attribute Utility Theory which requires that subjects employ a fully conscious exposition of all goals and attributes to find a proper weighting scheme. However, the practicality of this approach in a busy world where important decisions must be made rapidly without time for such exposition is an open question.Moreover, even with sufficient time to rely on a fully conscious exposition, decision makers are prone to implicit biases which may be difficult to recognize. Thus it is also difficult to render the decision process entirely conscious.
Acknowledging that both conscious and non-conscious processes are part of all decisions, some researchers have switched from focusing on whether decision makers should rely on either conscious deliberation or semi-conscious intuition to focusing on exploring the characteristics of decision contexts under which more intuitive ormore deliberate reasoning strategies lead to better outcomes. For example, Hogarth (2002) explains that intuitive processing is more beneficialwhen the information provided by the immediate context of the decision task is sufficient to make a valid judgment or decision. Hogarth points out a critical distinction between information that is merely presented by the decision context, and information that becomes internally represented for the decision maker. Deliberate reasoning is more beneficial when additional information, not initially represented internally, is required. An example of this distinction is the classic base-rate neglect problems in which base rates are presented, but do not become sufficiently represented in the subject’s reasoning process. Here, deliberate reasoning is necessary (but not sufficient) to bring base rate information into consideration.
Even when all the relevant information becomes represented internally, whether subjects utilize the information properly depends on factors such as whether the subject has had the opportunity to directly sample outcomes through experience with similar decisions, and whether past learning environments have provided veridical feedback for past responses.Although expert judgments are, in many domains, inferior to simple linear regression models, that their validity in some domains can be quite high, as in the predictive judgments of weather forecasters, is taken as evidence that good learning environments can produce good expert judges (Hogarth, 2002).
The emphasis onlearning experiencein intuitive processing is also found in cognitive neuroscience process models of decision making. In the influential “somatic marker” hypothesis, multiple competing responses are automatically induced for the decision maker by stimuli in the decision context, and in order to choose among these multiple, competing responses, the decision maker draws upon somatic markers (e.g. autonomic arousal components of visceral “gut feelings”) which represent the relative punishment or reward value of those responses relative to the stimuli (Naqvi, Shiv & Bechara, 2006). Punishment and reward values of stimulus-response pairings are acquired by experience, but their representation will only be veridical for the decision maker if feedback from decision maker’s past encounters has also been.
Cognitive neuroscience investigations of intuitive phenomena have raisedinterestingquestions about where intuitive feelings come from, what they represent, and how theycan influence executive-level processing that is central to decision making.For example, in a study of the neural mechanisms of dual system processing, the anterior cingulate cortex (ACC), which has been otherwise implicated in monitoring response conflict in complex cognitive tasks, has been shown to increase activity when the discrepancy between a normative and a likely descriptive response is widened suggesting that our brain has the capacity to detect an error, even if we are not consciously aware of the error (De Neys, Vartanian & Goel, 2008). Therefore, assuming the role ascribed to ACC activity is correct, a decision maker who is sufficiently sensitive to ACC activation could use that signal as an “intuitive” guide to select the normative response. Similarly, in a study where subjects were required to make intuitive “feeling of knowing” (FOK) judgments regarding the answer to math problems, event-related potentials (indicators of brain activity associated with a cognitive “event”) which corresponded with the subjects’ reported FOK occurred rapidly, before the subjects were consciously aware of their FOK and were more accurate predictors of whether the subject actually could provide the solution than the subject’s own verbalized judgment (Paynter, Reder & Kieffaber, 2009). Here again, there is evidence for non-conscious signals in the brain which are better calibrated than conscious, behavioral responses.
Evidence from both of the last two studies suggests that non-conscious brain activity is more accurately correlated with the “correct” response than the subject’s explicit behavior. The discrepancy between descriptive, explicit behavior and normative, implicit activity in the last two studies provides evidence contrary to the Panglossian position which is that no descriptive-normative gap exists. That experts in some domains can learn to make good use of intuition provides reason for hope that, in the interest of making us better decision makers, we can be taught to become more sensitive to the non-conscious signals of normativity.
References
Betsch, T. (2007). The nature of intuition and its neglect in research on judgment and decision making. In Plessner, H., Betsch, C., & Betsch, T. (Eds.), Intuition in Judgment and Decision Making (pp. 3-22). Mahwah, NJ: Lawrence Erlbaum.
De Neys, W., Vartanian, O., & Goel, V. (2008). Smarter than we think: When our brains detect that we are biased. Psychological Science, 19, 483-489.
Dijksterhuis, A., Bos, M. W., Nordgren, L. F., & van Baaren, R. B. (2006). On Making the Right Choice: The Deliberation-Without-Attention Effect. Science, 311, 1005.
Hogarth, R. (2002, October). Deciding analytically or trusting your intuition? The advantages and disadvantages of analytic and intuitive thought. ICREA and Pampeu Fabra University, Barcelona Spain. Retrieved April 4, 2009, from
Naqvi, N., Shiv, B., & Bechara, A. (2006). The role of emotion in decision making: A cognitive neuroscience perspective. Current Directions in Psychological Science, 15, 260-264.
Paynter, C. A., Reder, L. M., & Kieffaber, P. D. (2009). Knowing we know before we know: ERP correlates of initial feeling-of-knowing. Neuropsychologia, 47(3), 796-803.
Price, Mark C., & Norman, E. (2008). Intuitive decisions on the fringes of consciousness: Are the conscious and does it matter? Judgement and Decision Making, 3(1), 28-41.
Stanovich, K. E. (1999). Who Is Rational?: Studies of individual Differences in Reasoning (1st ed.). Mahwah, NJ: Lawrence Erlbaum.
Tversky, A., & Kahneman, D. (1974). Judgment under Uncertainty: Heuristics and Biases. Science, 185(4157), 1124-1131.
Wilson, T. D., Lisle, D. J., Schooler, J. W., Hodges, S. D., Klaaren, K. J., & LaFleur, S. J. (1993). Introspecting about Reasons can Reduce Post-Choice Satisfaction. Pers Soc Psychol Bull, 19(3), 331-339.