Lemonade, Willpower, and Expensive Rule-Based Analysis

This experiment used the attraction effect to test the hypothesis that ingestion of sugar can reduce reliance on intuitive, heuristic-based decision making. In the attraction effect, a difficult choice between two options is swayed by the presence of a seemingly irrelevant ‘‘decoy’’ option. We replicated this effect and the finding that the effect increases when people have depleted their mental resources performing a previous self-control task. Our hypothesis was based on the assumption that effortful processes require and consume relatively large amounts of glucose (brain fuel), and that this use of glucose is why people use heuristic strategies after exerting self-control. Before performing any tasks, some participants drank lemonade sweetened with sugar, which restores blood glucose, whereas others drank lemonade containing a sugar substitute. Only lemonade with sugar reduced the attraction effect. These results show one way in which the body (blood glucose) interacts with the mind (self-control and reliance on heuristics). The capacity for rational choice based on intelligent analysis and reasoning is one of the most remarkable and distinctive attributes of the human mind. But human decision making does not always adopt such high levels of analysis. Many of the choices people make are quick and effortless, and people’s decisions are often based on previously established heuristics, rather than a thorough application of the strict rules of logic. This inconsistency in human reasoning has been explained through dual-process accounts of decision making. According to these accounts, one reasoning process makes quick and automatic judgments based on associative and intuitive feedback, and the other process is more effortful and relies on the application of normative rules of reasoning. That judgment outcomes vary across contexts is said to be, in part, a reflection of these complementary systems. In the present research, we sought to establish psychological and physiological causes for the reliance on one decision-making process over the other. Specifically, we tested the hypothesis that more blood glucose (which serves as fuel for the brain) is needed for the effortful, rule-based process than for the less effortful process, so that the former type of decision making is impaired when glucose has been depleted by prior, even irrelevant, activities. For our test, we used a specific pattern of intuitive, heuristicbased decision making that has been identified in previous work. We sought to show (a) that the influence of this heuristic is increased when cognitive resources, presumably including blood glucose, have been depleted by prior acts of self-control and (b) that effortful processing can be increased (and the reliance on heuristics reduced) by administering a snack that restores blood glucose to its normal levels. TWO DECISION SYSTEMS Several accounts have analyzed human reasoning into two complementary processes. Epstein (1994) proposed that information processing is executed by an experiential system that is holistic, affective, and associationistic and by a rational system that is analytical, logical, and reason oriented. Sloman (1996) described two similar systems: an associative system based on automatic intuition and a rule-based system based on deliberation and the manipulation of symbols. More recently, Stanovich Address correspondence to E.J. Masicampo, Department of Psychology, Florida State University, Tallahassee, FL 32306-4301, e-mail: masicampo@psy.fsu.edu. PSYCHOLOGICAL SCIENCE Volume 19—Number 3 255 Copyright r 2008 Association for Psychological Science (1999; and later, Kahneman, 2003) described the two processes as a heuristicand association-based process, referred to as System 1, and a controlled, rule-based process, referred to as System 2. Common to all these models of reasoning is the idea that one of these systems (System 2) is more effortful and rule based, whereas the other (System 1) is relatively effortless and relies on quick associations and heuristics. Theories of reasoning differ in their descriptions of the relative adaptiveness of the two systems. For example, early work charting the limitations of human rationality described the heuristic-based System 1 process as both biased and prone to systematic error (Kahneman, Slovic, & Tversky, 1982). Furthermore, Kahneman (2003) proposed that one major purpose of the rule-based System 2 process is to monitor System 1 decisions and to correct them when they are in error. Alternative accounts have characterized System 1 processes as providing a fast and frugal substitute for expensive System 2 thinking (Gigerenzer & Goldstein, 1996), and, indeed, some evidence suggests that the role of intuitive, heuristic strategies is to usurp effortful, analytical processes over time, so that comparable benefits can be achieved with much less effort (Reyna & Ellis, 1994). The aim of the current study was not to compare the adaptive value of the two reasoning systems, but to examine a potential asymmetry in the demands that they exert. We proposed that the effortful System 2 process requires access to limited psychological and physiological resources, and we anticipated that when we manipulated the availability of these resources in a decision-making context, we would observe concordant changes in System 1 versus System 2 processing. THE ATTRACTION EFFECT Huber, Payne, and Puto (1982; also Simonson, 1989) identified a useful procedure for studying the different reasoning processes. In their study, some participants faced a difficult decision between two options that traded off on important, relevant dimensions. Other participants faced a choice between those same two options plus a third, ‘‘decoy’’ option. The decoy option resembled one of the others but was inferior to it in every respect. Logically, participants should have ruled out the decoy, which would have left them facing the same two options as in the basic, two-option version of the dilemma. Thus, their choices should have exactly paralleled those of the participants who had only those two options to begin with. The researchers found, however, that even though no one chose the decoy, participants came to favor whichever of the original options was similar to the decoy. In other words, the decoy exerted an irrational bias (attraction) on the two prime options. This phenomenon is called the attraction effect or the asymmetric dominance effect. Because the attraction effect exerts a seemingly irrational influence on the decision process, its impact on choice can be seen as working through the intuitive System 1 process rather than the rule-based System 2 process. Indeed, Dhar and Simonson (2003) concluded that the attraction effect operates primarily through intuitive and perceptual System 1 processes, and recent findings have suggested that when cognitive (System 2) resources are low, the attraction effect tends to be increased (Pocheptsova, Amir, Dhar, & Baumeister, 2007). Furthermore, work by Simonson (1989) suggests that the attraction effect serves as a likely tiebreaker when thorough, effortful analysis fails to produce a clear preference. Thus, the attraction effect is greatest when analytical System 2 processes fail and the decision-making process defers to System 1. DEPLETED RESOURCES AND BLOOD GLUCOSE Research on self-regulation and the self’s executive function has established that they rely on a limited resource. Early studies found that after people engaged in one act of self-control, their performance on subsequent and seemingly irrelevant self-control tasks was impaired (Baumeister, Bratslavsky, Muraven, & Tice, 1998; Muraven, Tice, & Baumeister, 1998). These findings suggested that some resource needed for optimal control on the second task had been depleted during the first task. The limited resource used for self-control is also used for effortful decision making. Vohs et al. (2007) showed that making multiple or difficult decisions led to subsequent decrements in self-control, as if the same vital resource was needed for both kinds of tasks. In an unpublished study, Pocheptsova et al. (2007) reversed this procedure to assess the effects of acts of self-control on subsequent decision making. They found that heuristic-based decision-making processes, including those that drive the attraction effect, were significantly stronger among depleted participants than among those who were not depleted. These findings support the argument that System 1 processes gain increased influence when scarce cognitive resources do not allow for optimal System 2 reasoning. What is the nature of the resource that is depleted? The lay term willpower has a long history, but has not generally been accorded much respect in psychological theorizing. Gailliot and Baumeister (2007) proposed that willpower is more than a metaphor and that blood glucose may be an important physiological basis for it. Glucose is fuel that is consumed to provide energy for all brain activities, but some brain processes consume much more energy than others. A series of studies by Gailliot et al. (2007) showed that blood glucose levels dropped when people performed laboratory tasks that required self-control (but not neutral tasks) and that these low levels of blood glucose were significantly correlated with poor performance on subsequent behavioral measures of selfcontrol. These studies also showed that replenishing blood glucose with a drink containing sugar counteracted the depletion and restored self-control performance to a good level, whereas sugar-free diet drinks had no such effect. 256 Volume 19—Number 3 Decision Making and Glucose