Sunteți pe pagina 1din 17

This article was downloaded by: [Consorci de Biblioteques Universitaries de Catalunya] On: 3 December 2010 Access details: Access

Details: [subscription number 789296667] Publisher Psychology Press Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 3741 Mortimer Street, London W1T 3JH, UK

Psychological Inquiry

Publication details, including instructions for authors and subscription information: http://www.informaworld.com/smpp/title~content=t775648164

Intuition: A Challenge for Psychological Research on Decision Making


Robin M. Hogartha a Department of Economics and Business, Universitat Pompeu Fabra, Barcelona, Spain Online publication date: 03 December 2010

To cite this Article Hogarth, Robin M.(2010) 'Intuition: A Challenge for Psychological Research on Decision Making',

Psychological Inquiry, 21: 4, 338 353 To link to this Article: DOI: 10.1080/1047840X.2010.520260 URL: http://dx.doi.org/10.1080/1047840X.2010.520260

PLEASE SCROLL DOWN FOR ARTICLE


Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf This article may be used for research, teaching and private study purposes. Any substantial or systematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material.

Psychological Inquiry, 21: 338353, 2010 Copyright C Taylor & Francis Group, LLC ISSN: 1047-840X print / 1532-7965 online DOI: 10.1080/1047840X.2010.520260

Intuition: A Challenge for Psychological Research on Decision Making


Robin M. Hogarth
Department of Economics and Business, Universitat Pompeu Fabra, Barcelona, Spain Intuition represents an enormous challenge for research on decision making. What is intuition? How does it modify our appreciation of cognitive abilities? When should people trust intuition? These questions set the agenda for this article, which (a) denes intuition, (b) comments on how intuition has been viewed across time in the decision making literature, (c) stresses the need to specify different types of intuition, (d) discusses when intuition is likely to lead to good decisions, and (e) presents four challenges. These are, rst, elucidating the evolution of preferences; second, illuminating culturally acquired values such as morals; third, the need to educate intuitive responses; and fourth, problems in using intuition for decision making in a changing world. However, the major challenge facing intuition research is the need for conceptual work to dene the nature and scope of different intuitive phenomena. To be useful, the concept should not become too broad.

Downloaded By: [Consorci de Biblioteques Universitaries de Catalunya] At: 12:45 3 December 2010

The study of decision making has attracted considerable scientic attention from many disciplines including psychology, economics, sociology, political science, biology, neuroscience, and different areas of business administration. There are at least two reasons. First, decision making provides an intrinsically interesting set of phenomena. Second, good description can lead to good prescription. If we know how people make decisions, we may be able to help them make better decisions. The study of decision making is, however, subject to a major difculty. Whereas we can typically observe the actions that people takeand often the ensuing outcomeswe cannot observe the actual processes used. As yet, we lack the technology to see into the mind,1 and thus must posit underlying models. Of course, we can always reject models if data are inconsistent with observations (e.g., of actions or process data such as response latencies). But the fact remains that investigators typically have to take leaps of faith concerning models of underlying processes be they formal analytic representations such as expected utility in economics or, say, more qualitative processes suggested by psychological theories. Whereas appeals to introspection do not provide scientic proof, most people resonate to the notion that there are differentespecially two different ways of apprehending experience and making decisions. Broadly speaking, we recognize that whereas some decisions can be taken into use Gladwells (2005) terma blink, others require us to think. In
1 This statement is made despite the recent and important advances in neuro-scientic technology.

addition, we often have a choice between experiencing the consequences of blinking or thinking. Many researchers too have postulated so-called dual process theories (Chaiken & Trope, 1999) that distinguish between what can be roughly called intuitive and analytic decision-making processes, although these go under different names, for example, System 1 and System 2 (Stanovich & West, 2000), experiential and rational (Epstein, 1994), or tacit and deliberate (Hogarth, 2001). Somewhat earlier, Hammond conceived of intuition and analysis as being the endpoints of a continuum of cognitive styles that can be matched to the requirements of a parallel continuum of different task characteristics (Hammond, Hamm, Grassia, & Pearson, 1987). Intuition plays an important role in lay psychological reasoning. Consider, for example, appeals that are made to womens intuition or stories of remarkable intuitive hunches made by scientists and business tycoons. In most cases, intuition here refers to the use of knowledge that cannot be made explicit but is surprisingif not a little magicalin its accuracy. A little less than a century ago, the psychoanalytic pioneer Carl Jung (1926) used the term in expounding his theory of personality and today tests based on Jungs work such as the MyersBriggs Type Indicator (Briggs & Myers, 1976) purport to measure preferences for intuitive thinking. (However, see also Epstein, Pacini, Denes-Raj, & Heier, 1996.) Recent years have seen a surge of interest in intuition as a scientic topic in psychology. This has been witnessed by several books (Gigerenzer, 2007; Gilovich, Grifn, & Kahneman, 2002; Gl ockner & Witteman, 2010b; Hogarth, 2001; Klein, 1998; 338

THE CHALLENGE OF INTUITION

Downloaded By: [Consorci de Biblioteques Universitaries de Catalunya] At: 12:45 3 December 2010

Myers, 2002; Plessner, Betsch, & Betsch, 2008; SadlerSmith, 2008) and articles (Hodgkinson, Langan-Fox, & Sadler-Smith, 2008; Kahneman & Klein, 2009) that would provide more than enough material for advanced graduate-level courses. In addition, work on intuition in the applied area of management has evolved from recounting anecdotes to a more scientic basis (see, e.g., Dane & Pratt, 2007, 2009; Khatri & Ng, 2000). In this article, I take stock of this interest and ask how it can add to our understanding of judgment and decision making. The article is organized as follows. First, I briey dene what I mean by intuition. Second, I comment on how intuition has been viewed across time in the judgment and decision-making literature. In particular, I note that recent work suggests a more optimistic view of human cognitive abilities. Third, I make the point that intuition covers many different functions and that these need to be identied to understand when it is and is not likely to be effective. However, this is complicated by the fact that the outputs of intuition have their origins in multiple information-processing systems that operate below the level of consciousness and are heavily dependent on tacit learning. Fourth, I comment on whether and when intuition or analysis is a better way to make decisions. This debate, however, is possibly misleading in that effective decision making requires intuition and analysis. Thus, it is essential to understand the relevant advantages and disadvantages of both. Next, I discuss four challenges posed by the role of intuition in human action and decision making. These concern, rst, the evolution of preferences; second, judgments of culturally induced values such as morals; third, the education of intuitive responses; and fourth, how to harness the benets of intuition in a changing world.2 I conclude by noting that, given the current stage of knowledge, it has been helpful to dene the scope of intuitive phenomena broadly. However, this incurs the danger of making the concept of intuition too inclusive. Thus, the major challenge facing intuition research is to classify intuitive phenomena in ways that can lead to more precise and useful questions.

striking than the differences. For this reason, I stick with the denition that I provided some 10 years ago, namely, the essence of intuition or intuitive responses is that they are reached with little apparent effort, and typically without conscious awareness. They involve little or no conscious deliberation (Hogarth, 2001, p. 14). To this I add that intuitive judgments are typically but not alwayscorrelated with speed and often a sense of condence. Moreover, they may have cognitive or emotional origins, that is, result from pattern recognition (see, e.g., Klein, 1993) or feelings such as, say, fear (see, e.g., Damasio, 1994; Le Doux, 1996). Most authors also agree that intuitions are experienced in a holistic manner (e.g., Sadler-Smith, 2008), that is, conclusions are based on overall impressions as opposed to cognitively combining separate elements of a problem. Finally, as I argue next, intuition is the result of learning. Two Other In- Words Intuition is often confused with two other words that also start with the prex in-specically, instinct and insight. An instinct is a reaction that is innate, for example, shutting ones eyelid automatically in response to a puff of cold air. It is important to distinguish the two concepts by the fact that, contrary to instincts, intuitions represent learned behavior. Indeed, as previously noted, learning and intuition are inseparable. Describing intuitive reactions of behavior as instinctive then should only be considered metaphorically (e.g., as in a tennis champions instinctive reactions in a game). Insight means seeing into the structure of a problem or issue. Clearly one can achieve insight into a problem in an intuitive manner. However, insight can also be achieved through deliberate processes. Imagine, for example, following the steps of a proof for a mathematical theorem. Insight and intuition are not interchangeable. Finally, cognition is complex. Most of the actions we take involve a mixture of intuitive and non-intuitive elements or processes (as discussed further next). In many cases, the characterization of actions as intuitive or non-intuitive is a matter of degree.

Dening Intuition Several investigators have summarized different denitions of intuition. For example, Dane and Pratt (2007, Table 1) provided denitions from 17 different authors. Whereas there are differences between these denitionssome authors weighting some aspects more than othersit is the similarity that is more
2 A further challenge that I do not consider here is to specify the advantages and disadvantages of different methodologies for research on intuition. For discussion of these issues, see Dane and Pratt (2009) and Gl ockner and Witteman (2010b).

Evolving Views in the Judgment and Decision-Making Literature Taking the mid-1950s as the starting point of this literature, it is clear that intuitive judgment has not been held in high esteem by researchers for most of the last 60 years. First, in a landmark study on the accuracy of intuitive or clinical judgment, Meehl (1954) documented 339

HOGARTH

that statistical rules could outpredict experts in certain diagnostic tasks. Of course, he did not establish that experts were actually relying on intuition (as previously dened) and so the real comparison was between statistical decision rules and unaided human judgment (assumed to be mainly intuitive). Other researchers have documented similar ndings (Grove, Zald, Lebow, Snitz, & Nelson, 2000). Second, analysis of the accuracy of human judgment in bootstrapping studieswhere predictions of models of peoples judgments are compared with the predictions people make themselveshas led to the conclusion that the intuitive component of judgment adds little to predictive performance.3 Moreover, it is more than compensated for by the consistency of models when these are used instead of humans for prediction (Camerer, 1981; Goldberg, 1970; Karelaia & Hogarth, 2008). Third, phenomena such as illusory correlation (Chapman & Chapman, 1967) established the notion that unaided human judgment (i.e., intuition) could be mistaken in systematic ways. Fourthand perhaps most inuentiallythe heuristics and biases program on probabilistic reasoning (Kahneman, Slovic, & Tversky, 1982; Tversky & Kahneman, 1974) contributed to the impression that, in many inferential tasks, intuitive judgments are systematically biased. Once again, intuition was not always explicitly dened, but there was an underlying assertion that peoples judgments were being systematically inuenced by forces beyond their conscious awareness (e.g., the implicit use of similarity in the representativeness heuristic). At one level this preoccupation with dysfunctional aspects of unaided human judgment (intuition) was healthy. It suggested that, in many cases, people should seek help in making judgments in applied areas such as management (Bazerman, 2005), probability assessment (Hogarth, 1975), or medicine (Groopman, 2007). However, in addition to researchers who aimed to show that inferences being made about human judgment abilities were incorrect (e.g., Cohen, 1981), others questioned the boundary conditions of the heuristics and biases movement (see, e.g., Hogarth, 1981). From the perspective of this article, important contributions were made by Hammond, Gigerenzer, and Klein and their collaborators. Building on Brunswiks (1952) theory of probabilistic functionalism, Hammond developed the cognitive
intuitive component in judgment is analyzed using residual terms in Brunswiks (1952) lens model framework (Karelaia & Hogarth, 2008). However, there is a legitimate argument that much of the systematic, linear processing captured within the lens model framework could itself be intuitive in nature. Thus, limiting the contribution of intuition merely to correlations involving residuals underestimates the role of intuition in lens model-type situations that can involve automatic processing (Gl ockner & Betsch, 2008b).
3 The

Downloaded By: [Consorci de Biblioteques Universitaries de Catalunya] At: 12:45 3 December 2010

continuum theory mentioned briey earlier. The key idea is that both styles of judgment and task characteristics can be arranged along parallel continua ranging, in the case of judgment, from intuition to analysis, and in the case of tasks, from intuition inducing to analysis inducing. The key to good judgment is to match task demands and cognitive style, and in an important empirical investigation concerning judgments of highway safety, there was considerable support for this framework (Hammond et al., 1987).4 The contribution of Gigerenzer was to ask two questions similar to those posed by Hammond. First, what kinds of tasks is the human mind naturally suited to handling? Second, how do the characteristics of tasks people face in laboratory situations match those encountered in their natural environments or ecology? These questions were answered in two separate investigations. In one, Gigerenzer and Hoffrage (1995) showed that reframing Bayesian updating problems using so-called natural frequencies induced far more accurate responses by experimental participants.5 Second, overcondence in judgment can be considerably mitigated when questions posed by investigators involve samples from the domain of participants natural environments (Gigerenzer, Hoffrage, & Kleinb olting, 1991). Finally, the work of Klein and his collaborators conducted within the naturalistic decision-making paradigm is relevant because of its emphasis on how experts use intuitive pattern recognition in their professional judgments (Klein, Orasanu, Calderwood, & Zsambok, 1993). I consider the importance of recognition further next. Changes in our Appreciation of Cognitive Abilities Research on decision making has also been greatly inuenced by Simons (1955) concept of bounded rationality. The key notion is that, because people are physically incapable of performing all the rational calculations required by economic theory, they resort instead to other mechanisms. For example, in his original article Simon (1955) suggested that people reduce cognitive load by using aspiration levels. That is, instead of maximizing by identifying the best alternative in a large choice set, people adopt the simpler and
4 Parenthetically, in an intriguing earlier study, Peters, Hammond, and Summers (1974) illustrated how using analytic and intuitive processes leads to different types of error distributions in judgmental tasks. Thus, determining whether analysis or intuition is better also requires considering what kinds of error are worse and by how much. 5 Representation by natural frequencies incorporates base rates in the data presented in a way that is natural for people to consider in their calculations. Thus, base rate information is not ignored. For an even simpler and natural characterization of probabilistic reasoning, see Hogarth and Soyer (2010).

340

THE CHALLENGE OF INTUITION

practical goal of nding an alternative that is good enough (i.e., that meets an aspiration level). Clearly, this process is not rational in that it cannot guarantee identifying the best alternative. However, people are typically content to satisce. Simons ideas went a long way to help explain why people resort to heuristics or simplied decision rules that, although implying systematic errors in some situations, are generally valid in their use (Gigerenzer, Todd, & the ABC Group, 1999; Kahneman et al., 1982). Whereas scholars have alternatively emphasized the functional and dysfunctional aspects of these heuristics (see, e.g., Klayman & Ha, 1987), the fact remains that our intellectual legacy is a collection of mechanisms (or heuristics) that function in the manner of intuitive rules. That is, when faced with specic tasks, people intuitively (i.e., without conscious awareness of the process) make judgments that experiments have shown to be consistent with specic heuristics (cf. Kahneman & Frederick, 2002).6 To the extent that these heuristics represent intuitive judgment, it can be argued that intuition is awed (when the emphasis is on errors as in Gilovich et al., 2002) or highly effective relative to other decision rules (as in Gigerenzer et al., 1999, although heuristics advocated in this latter approach often have a strong deliberative component). Of course, if decision rules do not maximize then, by denition, they cannot be accurate in all environments. It is therefore important to determine how characteristics of rules match the demands of environmental tasks and consequently when particular rulesor intuitionsare likely to be effective (Hogarth & Karelaia, 2007). Simons (1955) work was entirely consistent with the Zeitgeist of the cognitive revolution thatfrom an information processing viewpointdepicted the mind as a computer with impoverished computational power. However, the interesting implication of work on intuition is that it has highlighted dimensions of information processing where cognitive limitations do not play an important role. One of these practical nonlimitations is the human ability to process and encode frequency information accurately (see also next). The evidence on this topic is both uncontroversial and overwhelming (Hasher & Zacks, 1979, 1984; Zacks & Hasher, 2002). That it is a basic cognitive mechanism that was probably developed through evolutionary pressures is reinforced by the ndings that several nonhuman species show similar capacities, for example, in understanding frequency distributions associated with different sources of food. It is clearly an important intuitive asset.
6 I do not mean to imply here that all heuristics are intuitive rules. In many cases, people will deliberately choose to use heuristics instead of attempting more complete and rational analysis of problems. For example, consider Simons (1955) satiscing rule.

Downloaded By: [Consorci de Biblioteques Universitaries de Catalunya] At: 12:45 3 December 2010

A second important intuitive asset is the capacity for recognition. This literature has been well summarized by Goldstein and Gigerenzer (2002), who noted that because recognition continues to operate even under adverse circumstances, and it can be viewed independently from the rest of memory, we view it as a primordial psychological mechanism (p. 77). They went on to cite several studies that attest to just how remarkable human memory is in its ability to recognize previously encountered stimuli. Indeed, recognition memory is at the heart of intuitive expertise, such as the ability of chess masters to recognize many thousands of positions from possible games (Ericsson & Charness, 1994; Klein, 1993; Simon & Chase, 1973). A third factor is the growing evidence of the importance of automatic processing. In particular, this has been shown to occur even in decision-making tasks that are subject to severe cognitive limitations such as memory and time constraints where researchers previously assumed that people would have relied on deliberate heuristic decision rules (see, e.g., Dougherty, Gettys, & Ogden, 1999; Gl ockner & Betsch, 2008a, 2008b). Finally, recent research shows that even limitations on attention may be less severe than previously thought. For example, there is evidence that the visual system can extract categorical informationsuch as the presence of cars or animalsin scenes after very brief exposure and even when performing unrelated tasks that demand attention (Peelen, Fei-Fei, & Kastner, 2010).

Functions and Varieties of Intuition Intuition is clearly involved in many types of cognitive operations. As such, it is important that these be classied in ways that facilitate understanding. One important distinction lies in the content of intuitions. Intuitions can be thought of as inferences that look forward or backward in time or, alternatively, they represent a stock of knowledge that I have referred to as cultural capital (Hogarth, 2001). Forward-looking inferences are essentially predictions, for example, estimating where a ball will land after being hit in the air. Backward-looking inferences are diagnostic in nature. Something has happened and the outcome of intuition is a hypothesis that could explain why. Bowers, Regehr, Balthazard, and Parker (1990) referred to the distinction between intuition in the context of justication (forward-looking predictions) and the context of discovery (backward diagnosis). Dane and Pratt (2009) also provided a classication of types of intuition. They distinguished problem solving (based on pattern recognition), moral (a subset of the cultural capital idea discussed further next), and creative (described as feelings that arise when knowledge is combined in novel 341

HOGARTH

Downloaded By: [Consorci de Biblioteques Universitaries de Catalunya] At: 12:45 3 December 2010

ways). The latter, it can be assumed, is similar to backward diagnosis. The notion of intuition as a stock of knowledge or cultural capital lies in the fact that, over time, we acquire much tacit knowledge on which we draw in making all kinds of inferences be they professional or private (e.g., our preferences, job-related expertise, moral judgments, or cultural reactions to events). And, of course, forward and backward inferences are not independent of our stocks of knowledge (or cultural capital) but draw on them. However, the particular intuitions that emerge may depend on whether inferences are forward or backward looking in time. Much attention has been paid as to whether and when intuitions are correct or more accurate than the outcomes of deliberate processes. This is an important issue that I consider throughout this article. However, if we consider the different types of intuitions just dened (forward and backward inferences and stocks), there is an important, conceptual problem. This is that although, in principle, the accuracy of particular forward or backward inferences can be veried, this is not true of intuitions from all stocks of cultural capital. Is it correct, for example, to have a specic, intuitive, cultural reaction toward, say, veils worn by Muslim women? Intuitive reactions to this practice clearly vary by culture, but it is not clear what is correct or even why. An important implication of studying intuition is that it demands the simultaneous consideration of many phenomena that psychologists typically study in relative isolation (Hodgkinson et al., 2008). Not only does intuition engage research on cognitive processes, it also demands that researchers adopt a broad perspective and deal with issues of nonconscious thought processes, learning, affect, and the integration of different information-processing systems. Many Information-Processing Systems Earlier I referred to the extensive work conducted within the conceptual framework of dual process models (Chaiken & Trope, 1999). As noted, this provides a useful distinction between judgments or actions that can be thought of as being intuitive or analytic in nature and has helped to illuminate many phenomena (Epstein, 1994). Yet, if one pushes the dual process idea further and, in particular, starts to look deeper into nonconscious and automatic processing, it becomes clearer that humans have many different informationprocessing systems and that there are further useful distinctions to be made within the two processes of dual models. In a stimulating article, Gl ockner and Witteman (2010a) made precisely this argument and suggested that intuitive processes might be usefully categorized in four different ways. Two of these, associative in342

tuition and matching intuition, involve processes of learning and retrieval. The rst deals primarily with simple stimulus-response type processes such as classical conditioning, social learning, and so on. This thus includes phenomena such as the automatic encoding of frequency information. The second (i.e., matching intuition) involves the learning of exemplars, prototypes, images, and schemas and retrieval processes that depend on the matching of stimuli to exemplars and prototypes. As such, this is the basis of much professional expertise that is based on recognition. The third and fourth categories, accumulative intuition and constructive intuition, are concerned mainly with the integration of information from memory traces and/or currently perceived information. They are distinctive in the senses suggested by their names. Accumulative intuition deals with the accumulation of evidence from different sources, whereas constructive intuition involves the activation of related information and the construction of consistent mental representations. Gl ockner and Witteman (2010a) correctly pointed out that these four categories of intuition are not completely distinct but that because they capture somewhat different phenomena, it pays to make the distinction. For example, whereas affect can be involved in the rst three types of intuition as an input, it can be both an input and an output in constructive intuition. The usefulness of the particular categories proposed by Gl ockner and Witteman (2010a) is an open question. However, their work is welcome in that it suggests ways of moving beyond simplistic explanations that depend on dual-process classications such as System 1 and System 2. There is a compelling need to understand better the varieties of System 1 behavior. Building on the work of Reber (1989, 1993), in Hogarth (2001) I considered discriminating between three levels of automatic or subconscious processing and asking about the extent to which people should trust the outcomes of these different levels of processes. The levels were labeled basic, primitive, and sophisticated. The basic level refers to instinctive or biological reactions such as when we sense that we are hungry or thirsty. The primitive level is involved in a variety of basic information-processing functions that are more or less devoid of meaning, affect, or interpretation (Reber, 1989, p. 231). This thus includes, for example, the automatic encoding of frequency information, and so on, that we share with lower animals and that operates in a machine-like manner. In contrast, the sophisticated unconscious is quintessentially human and involves the interpretationthrough meaning and affectof experience (see also Lewicki, Hill, & Czyzewska, 1992). If we ask about the inherent wisdom of the subconscious at these three levels, we might reach different answers. At the basic level, our responses have been tuned by natural selection and thus should generally

THE CHALLENGE OF INTUITION

Downloaded By: [Consorci de Biblioteques Universitaries de Catalunya] At: 12:45 3 December 2010

be trusted. However, there are exceptions. For example, not all reactions to eat when hungry are functional when one considers how the supply of food (in terms of both quantity and variety) differs today from that experienced by our ancestors. At the primitive level, it is clear that one can trust that the process has faithfully recorded ones experience. However, whether the experience is relevant to the situation currently faced has to be assessed (see the upcoming discussion in the Learning section). As to the sophisticated level, the wisdom is far harder to assess and I suspect is subject to greater variation (on some scale of bad to good) than at the other levels. Another way of putting this is to say that intuitions from the sophisticated level of the subconscious provide signals that require interpretation (see also Damasios, 1994, somatic marker hypothesis). Hence, it is important that people learn to control their intuitive reactions through deliberate analytic thoughts or actions. For example, it may be impossible not to have certain intuitive social prejudices. But that this does not mean that one should be excused for acting on them. In short, an important challenge for intuition researchers is to identify useful ways of classifying different mechanisms of subconscious informationprocessing systems and, by specifying their functions, to illuminate the conditions under which they are functional and dysfunctional for the organism (see also Bargh & Morsella, 2008). Learning There is almost universal agreement that, even though some responses people make are instinctive (see earlier), intuition is shaped by learning.7 In developing some intuitions, we may in fact be aware of the learning process but, with the passage of time, the reactions acquired are likely to be automated such that one is no longer aware of the fact that they are being reinforced and adapted by experience. For the most part, the learning processes involved in the honing of our intuitions escape conscious attention. There is extensive evidence, for example, that we maintain frequency counts of events that occur in our environments even when we are not specically motivated to do so (e.g., number of times we have observed certain people in a street) and can even break these down into categories if asked more rened questions (e.g., what proportion were women?). Moreover, this skill appears to be quite accurate across the life cycle and hardly varies as a function of individual differences (Hasher & Zacks, 1979, 1984; Zacks & Hasher, 2002; see also earlier). In addition, the work by Reber (1989, 1993) on im7 That we are by nature prepared to learn some reactions more than others (see, e.g., Seligman, 1970) should not distract from this assertion.

plicit grammars shows that people can learn new rules to relate words even when involved in tasks which demand active attention on other issues. Indeed, the human ability to apply grammatically correct rules in speechbut without being able to articulate them is a stunning everyday example of the power of tacit learning to form accurate intuitions. In short, our naive human learning processes are remarkably well adapted to the regularities of the environments that we encounter. From the viewpoint of establishing valid intuitions (in a predictive sense), this is both a great strength and a source of weakness. To highlight this, in Hogarth (2001) I made a distinction between what I called kind and wicked learning environments. I dened kind learning environments as those where the information tacitly processed leads to valid inferences, for example, when the sample of instances the person has encountered is representative of the environment in which the ensuing intuitive judgment is applied. Feedback is neither missing nor distorted, and so on. On the contrary, in wicked learning environments, samples of experience are not representative and feedback might be missing or distorted. Worse still, mistaken beliefs can lead to dysfunctional actions in the form of self-fullling prophecies (Einhorn & Hogarth, 1978). Examples of effects of wicked environments have been documented by several researchers and used to explain several different types of cognitive biases such as illusory correlations (Denrell, 2005; Fiedler, 2000; Fiedler & Juslin, 2006). The point made by these studies is that, whereas people might process the data they see in an appropriate manner, they lack the metacognitive ability to correct for sampling biases and/or missing feedback (however, see Elwin, Juslin, Olsson, & Enkvist, 2007). How to teach people to develop such skills is, in my view, an important issue for research (see also next). Parenthetically, in a recent theoretical analysis, Le Mens and Denrell (2010) pointed out how, even if samples are representative of underlying processes, people can still acquire biased beliefs if the selections of the samples they see reect differential payoffs. The concepts of kind and wicked environment lead naturally to describing when intuitions can and cannot be trusted (in a predictive sense). When a persons past experience is both representative of the situation relevant to the decision and supported by much valid feedback, trust the intuition; when it is not, be careful (cf. Gl ockner & Witteman, 2010a; Kahneman & Klein, 2009). Unfortunately, in many cases, it is unclear how representative the situation is. Moreover, if people discount uncertainty by, for example, relying on the representativeness heuristic (Tversky & Kahneman, 1974), the accuracy of their judgment will be inferior to that of an explicit analytical ruleprediction by the relevant base rate. 343

HOGARTH

This discussion of learning has been framed so far as though it only applied to specic cognitive acts. However, it should be clear that it also relevant to considerations of affect and complex stimuli like patterns. Thus, the somatic marker hypothesis proposed by Damasio (1994) relates to the automatic learning of affective reactions to different types of stimuli that signal appropriate or inappropriate choice alternatives. Similarly, chess masters, who have experienced many thousands of different situations, demonstrate effective pattern recognition skills when confronted with new challenges (see, e.g., Chase & Simon, 1973).
Downloaded By: [Consorci de Biblioteques Universitaries de Catalunya] At: 12:45 3 December 2010

Emotional Reactions In many ways, intuition is experienced as a feeling of knowing, and many authors have explicitly discussed the role of the emotional system in intuitive judgments (see, e.g., Damasio, 1994; Le Doux, 1996; Zajonc, 1980). In the decision-making literature, there has been a growing appreciation of how emotional states affect judgments of risk and probability (see, e.g., Isen, 1993; Loewenstein, Weber, Hsee, & Welch, 2001). In some cases, intuitive judgments are simply emotional reactions (cf. Slovic, Finucane, Peters, & MacGregor, 2002); in other cases, emotion acts as intuitive moderators that affect judgments in a partial manner (Hogarth, Portell, & Cuxart, 2007; Hogarth, Portell, Cuxart, & Kolev, in press). Of interest, subtle priming of emotions can have important effects on judgmental outputs (De Vries, Holland, & Witteman, 2008a, 2008b; Hsee & Rottenstreich, 2004). Clearly, emotion-affect-feelings are all central to the topic of intuition.

Blink, Think, or Smink? One source of the considerable lay interest in intuition lies in the recounting of stories where successful decisions in different walks of life have been attributed to effortless, intuitive reactions. For many, there is something emotionally satisfying about the notion that an instantaneous judgmentor blink (Gladwell, 2005)can be more correct (or accurate) than the outcome of a deliberative, rational analysis of a decision problem (Kahneman & Klein, 2009). From a scientic perspective, of course, there are at least two issues with interpreting such events. The rst is a disproportional recall of incidents where intuition was correct as opposed to incorrect. The true success rate of intuition is not known. The second is the unacknowledged importance of lucksometimes an intuitive response does get it right in difcult circumstances, for example, when forecasting stock prices. Despite a natural scientic skepticism about the relative success of intuitive reactions as opposed to 344

the outcomes of rational deliberative processes, it is nonetheless appropriate to ask when people would be better off trusting intuition (i.e., blinking) or analysis (i.e., thinking) in making decisions. One way to conceptualize this issue is to consider the different types of errors that can be made in intuitive as opposed to analytic thought and to consider how the relative sizes of the errors are likely to vary by the type of situation considered (Hogarth, 2005). As previously noted, intuitive judgments tend to be holistic and are based on reactions to specic cues (e.g., by recognition). In analytic judgments, on the other hand, the elements of a problem are identied and subsequently aggregated in some manner. To illustrate, consider the task of estimating ones grocery bill at the checkout aisle of a supermarket. First, note that the system used by the supermarket is entirely analytic. Each item has a unique price, and the bill is determined by explicitly adding the prices of all the products selected. The process is also quite transparent. This system can, however, involve errors. Items could be scanned incorrectly into the system and, if some items are entered manually, mistakes could be made by, say, inverting gures (e.g., 78.16 is entered as 87.16, etc.). But in this kind of situation, the probability of an analytic error is small.8 Now, despite the fact that your bill is estimated analytically, prior to receiving it, you probably already have an intuitive expectation as to roughly what it will be. You know, for example, how much you usually pay for groceries and approximately how much your current purchases differ from average. This intuitive judgment is made holistically and with little conscious effort. Indeed, you may not even be aware that you made the judgment unless the total announced by the cashier differs signicantly from the intuitive estimation. At the same time, although you know that your intuitive estimate is not exactly right, you expect it to be approximately correct. In short, errors in analytic thought are characterized by data entry problems and/or the inappropriate specication/application of a formula. Errors in intuitive thought are essentially those of bias induced from past experience (e.g., using an inappropriate anchor in judgment) or idiosyncratic momentary inuences such as the salience of some information or affective states (Gl ockner & Betsch, 2008a). In considering, therefore, whether analysis or intuition is likely to be more accurate, the relative magnitude of the different possible errors needs to be assessed across different types of situations. In Hogarth (2005), I suggested that task complexity is an important source of error in analytic thought. For example, whereas few people make errors in simple arithmetic tasks (e.g., 23 + 18 = ?), the same
8 Most people seem to accept that very few errors are made in that few ever actually check their bills.

THE CHALLENGE OF INTUITION

respondents could err in attempting to revise probabilities using Bayes rule. Thus, the probability of making errors in analytic thought can be thought of as an increasing function of task complexity (as measured by, e.g., number of variables, types of functions, weighting schemes, etc.). Moreover, the probability of making errors also reects individual differences, expertise in particular. Statisticians, for example, would undoubtedly make fewer errors in Bayesian probability revision tasks than members of the general public. It is less clear how to assess the probability of bias in intuitive thought except to say that this will reect the conditions under which specic responses have been learned, that is, were they acquired in kind or wicked learning environments? In addition, because tasks vary on how much they are or are not compatible with natural response modes, bias in intuitive responses can depend on how questions are asked. Consider, for example, judging visually the likely trajectory of a moving object as opposed to answering in response to a verbal description of the same problem (McCloskey, 1983). In summary, resolving the conict between intuition and analysis is easy when tasks are analytically simple. Stick to analysis.9 As tasks become more analytically complex, the advantage shifts to intuition, but this is subject to the provision that the decision makers intuitions have been honed in kind environments. This, for example, would characterize the behavior of chess masters whobased on extensive past experience with feedbackcan quickly identify what actions to take as soon as they see situations on a chess board.10 Moreover, the feedback they receive after making decisions is typically quite clear. A further point is that in many applied settings there is no obvious analytical model to which decision makers can appeal, and thus intuition is almost the only source of reaction. Consider, for example, certain kinds of aesthetic judgments or even a business person evaluating opportunities for an entirely novel product. Several studies have examined the differential effects on decision quality of requiring people to make reasons for choice explicit as opposed to trusting to immediate feelings. Results suggest that, if unsure about their preferences, deliberation can have distorting effects that might not occur if people were more explicitly aware of their preferences (Wilson, Kraft, & Dunn, 1989; Wilson & Schooler, 1991). In addition, a mismatch between the demands of the task and the cogni9 As a simple example, consider the M uller-Lyer illusion. Intuitively, we see (judge) the two lines to differ in length. However, simple analysis using a ruler is sufcient to resolve any conict that they are in fact the same length even though one might still see the lines as different. 10 Chess masters also check their intuitive responses by analytic considerations. However, their rst reactions might still limit the alternatives they generate (Bilali, McLeod, & Gobet, 2010.)

tive strategy used can be dysfunctional as demonstrated by Hammond et al. (1987) when engineers made judgments of the aesthetic quality of highways using analytic processes. Nonetheless, it has been found that decision aids that force people to be explicit about their reasons for decisions heighten satisfaction in choice relative to control groups without such aids (Kmett, Arkes, & Jones, 1999). The role of intuition in the form of unconscious thought has been highlighted by a series of studies by Dijksterhuis and his colleagues (see, e.g., Dijksterhuis, Bos, Nordgren, & van Baaren, 2006; Dijksterhuis & Nordgren, 2006). In their paradigm, participants are interrupted in the process of choosing and asked to defer their decisions until they have completed another experimental task. The surprising result is that these delayed choices turn out to be better than the choices made by control participants who are not subject to the same manipulation. The argument made is that the process of unconscious thought during the interruption stage somehow leads to better decisions. (Note that an essential part of this manipulation is that participants know they will be asked to make a decision later on.) These ndings can be interpreted as providing a rationale for delaying decisions by sleeping on them thereby allowing unconscious thought (intuition?) to improve them. The ndings and boundary conditions of Dijksterhuiss paradigm have, however, been questioned by other researchers (Acker, 2008; Payne, Samper, Bettman, & Luce, 2008; Newell, Wong, Cheung, & Rakow, 2009). Although intriguing, the issues are far from being settled. Elsewhere in this article, I have outlined different types of intuition and pointed out that not all can be empirically validated. Sticking to predictive intuitions, Kahneman and Klein (2009) emphasized an obvious but important point (see also Brunswik, 1952; Hammond, 1996). This is that predictive abilitywhether intuitive or analytichas an upper bound dened by the extent to which the relevant event can be predicted in the environment. If environmental predictability is low, the difference between intuitive and analytic thought might be moot although empirical studies show a slight advantage for analytic methods (Hogarth, in press). But blinking (intuition) and thinking (analysis) are not the only alternatives for prediction. There is also sminkinga term introduced by Makridakis, Hogarth, and Gaba (2009)to cover situations where human judgment is explicitly replaced by simple statistical models such as those banks use for credit scoring (i.e., a deliberate process designed to overcome the inconsistencies of thinking). These models (often based on heuristic principles, Gigerenzer et al., 1999) may not be optimal from an ex ante statistical viewpoint but have been shown, time and time again, to be effective ex post (Hogarth, in press). Whether people do or do 345

Downloaded By: [Consorci de Biblioteques Universitaries de Catalunya] At: 12:45 3 December 2010

HOGARTH

Downloaded By: [Consorci de Biblioteques Universitaries de Catalunya] At: 12:45 3 December 2010

not use such models can be conceived as involving an important conict between intuitive and analytic arguments. The intuitive belief is that it is important to maintain control over ones predictions (Kahneman & Klein, 2009). The analysis involves the realization that, in a noisy environment, giving up control to a simple model in fact provides more control over outcomes in that it increases the overall level of predictive ability. However to do this, one has to accept error to make less error (Einhorn, 1986), an analytical principle that confounds popular intuition. Finally, there is no simple answer as to whether intuition or analysis leads to better decisions. Clearly, there are some situations where the answer is evident (i.e., analysis for problems involving arithmetic and intuition when the complexity of analysis is overwhelming as in some kinds of aesthetic judgments). However, because most decisions involve both analytic and intuitive elements, I believe that the debate is better framed by asking how to bring both to bear on the issues. For example, previously I cited the example of chess masters who explicitly verify their intuitive reactions by analytical considerations. In Hogarth (2001), I advocated augmenting analysis of problems by considering the impact of ones feelings, that is, to explicitly treat emotions as data. Identifying ways in which people can successfully incorporate both intuition and analysis in decision making is an important challenge for future research (for an example, see Blattberg & Hoch, 1990).

Important Challenges As previously noted, the study of intuition requires multiple perspectives from psychologists. In this nal section, I raise four issues that, in my opinion, deserve special attention. The rst two involve judgments based on peoples stocks of intuitions or cultural capital. These concern (a) preferences and (b) judgments of culturally induced concepts such as morals. The third issue relates to whether and how intuitions can be educated, and the fourth raises the point that, if intuitions arise through experience, how can people and society as a wholeacquire intuitions relevant to a world that none of us has experienced? Preferences Normative theories of decision making distinguish clearly between preferences and beliefs (or inferences). Preferences capture how much people value or like the possible outcomes of decisions, whereas beliefs are made operational by the probabilities attached to relevant uncertain events. Moreover, in decision making, preferences and beliefs should be independent of each other. Whereas preference and inference are the 346

twin concepts around which much decision research is conducted, there is a large asymmetry in the type of attention paid to each. On one hand, theories of beliefs in the form of probabilities have been well conceptualized. Beliefs, it is argued, result from knowledge or data that people acquire and can be modied by processes of learning. Indeed, Bayes rule has been proposed as an optimal learning model. For example, if two people have quite different beliefs but then have access to the same relevant facts, they will both revise their beliefs, which with enough data will eventually converge. This is not to say, of course, that beliefs always follow Bayes rule (indeed there is a huge literature devoted to this topic, see, e.g., Gilovich et al., 2002). The important point is that there is a model that makes beliefs operational and species how they should be changed in the light of new evidence. The status of preferences, on the other hand, could hardly be more different. In economics, for example, preferences are assumed to be exogenous and unchanging (Stigler & Becker, 1977). In psychology, although it is recognized that preferences can change through experience, there are no normative models that suggest how preferences should change as a result of experience, that is, a model to parallel Bayes rule for inference (although one could argue that learning models can capture how preferences do change). Whereas the existence of a normative standard (i.e., probability theory) has been crucial to investigations of statistical intuitions or beliefs, the lack of such a model has undoubtedly handicapped research on preferences. How do people acquire preferences? How do these change through experience? To what extent are preferences the outcomes of conscious processes or do they result entirely from nonconscious processes and, as such, are in the domain of intuition? To date, answers to these important questions are quite incomplete. First, the possibility that preferences mightfor the most partbe acquired through tacit experience was highlighted by Zajoncs (1968) ndings of the mere exposure effect. In situations where people encounter stimuli that are not experienced as doing them harm, increasing positive affect is accompanied by increased exposure. Contrary to the popular saying that familiarity breeds contempt, it appears that it leads to liking (although there are boundary conditions such as states of satiation where one can have too much of a good thing). The important point raised by Zajonc is to suggest that our preferences are being constantly updated by experience even though we are unaware of the extent to which this happens. From a decision-theoretic standpoint, this is disturbing. It suggests that the preferences used to select outcomes will in turn be affected by how those outcomes are experienced so that choice becomes, in part, a self-fullling process. As an everyday example, imagine you have a preference for food

THE CHALLENGE OF INTUITION

Downloaded By: [Consorci de Biblioteques Universitaries de Catalunya] At: 12:45 3 December 2010

X over food Y. Then the choice of X could lead you to like it even more thereby reducing the chance that you will try Y in the future. As Zajonc (1980) elegantly argued, preferences need no inferences in that many objects (or situations) are immediately classied on a liking scale without conscious effort. Indeed, this and similar observations led Slovic et al. (2002) to suggest an affect heuristic by which people express preference for alternatives that they just like without being able to specify why. Indeed, the claim made is that an important part of our intuitive emotional apparatus is to help us make choices without the need for conscious effort, and this makes a lot of sense from an evolutionary viewpoint. In fact, it could be argued that choosing immediately what you prefer (like) may be a good choice strategy in the sense of satisfying ones true preferences. Consider, for example, situations where holistic appreciations of choice alternatives are needed such as when buying items of clothing. Whether one looks good in that new outt can typically be decided instantly (even though it is still prudent to check out rst impressions as well as the cost of the clothes!). Finally, the notion that we possess an automatic system for adapting our preferences to our experiences can be seen from two perspectives. One is that of cognitive efciency. The other is that, in many instances, our preferences are formed passively by our idiosyncratic experiences and thus might not be consistent with those we would choose after reection. Cultural Capital In Hogarth (2001), I introduced the notion that, across life, people acquire stocks of intuitions or cultural capital that are predominantly the product of tacit learning that has taught us how to survive in our particular cultural niches. Thus, people reared in different environments react differently to the same stimulus depending on their experiences. Consider cases as benign as the use of specic words by American or British speakers of English or as emotionally charged as objects that serve as symbols of different religions. The point about cultural capital is similar to preferences. People do not set out to choose their cultural capital in some objective manner. Instead, it represents the idiosyncratic nature of their experiences with life that were not all actively chosen. Thus, attitudes (including prejudices), reactions in social environments, and even religious beliefs are all largely the fruit of tacit learning. Interestingly, whereas individuals may not be aware of the extent to which they have been formed to think in this way, many organizations in society are aware of the power of tacit learning. Consider, for example, how religious, military, corporate, and academic institutions inculcate beliefs among their adherents. Most social learning results from absorbing

information about how to behave through implicit messages. Of interest, consultants interested in promoting new cultures in organizations explicitly argue that leaders should not tell people what to do. Instead, people should be given the opportunity to absorb a new culture through observing living examples that may be amplied by appropriate myths and stories (Schein, 1985). One interesting recent line of research has involved the role of intuition in moral judgments (see, e.g., Haidt, 2001, 2010; Narvaez, 2010). There are two important issues. First, what are the foundations of moral judgments? Second, what should be the relative roles in ethics of principles suggested by intuitive moral judgments, on one hand, and rational arguments, on the other (Haidt & Joseph, 2004)? Which takes precedence? The second question assumes an answer to the rst. Namely, through educationin the broadest sense including tacit learningpeople acquire moral intuitions that guide what is assumed to distinguish right from wrong. Thus the type of environment in which people are reared clearly plays a huge role in the development of moral intuitions (Narvaez, 2006). Consider, for instance, the contrast between rearing children in environments where violence is or is not tolerated. Moral intuitions are, thus, necessarily culturally and time specic. As an example, consider how many highly moral people accepted the practice of slavery through the 18th century and yet, today, their descendants would consider this abhorrent. Clearly, like many emotional reactions, moral judgments just happen. But the cultural capital of our moral intuitions is not immutable because, across generations, rationally derived arguments can affect morality and, by inuencing behavior, eventually change the moral intuitions of succeeding generations. In other words, old slave owners may never acquire a moral intuition that slavery is abhorrent, but their descendants (reared in a world where slavery is outlawed) acquire an intuitive moral sense that it is unacceptable. Intuitions involved in what I call cultural capital are far ranging. Here I have considered preferences and, briey, moral judgments that clearly merit much more elaboration. However, my main point is that through the development of cultural capital people acquire notions of many important concepts that could well be illuminated by recognizing their intuitive origins. For further examples, consider judgments of fairness, equity, social justice, and various forms of discrimination. One implication is that if you dont like the world the way it is now, you have to do something (perhaps by legislation) so that succeeding generations can acquire appropriate intuitions by experiencing a new status quo. Parenthetically, as a nal point specically related to decision making, consider how people assess the 347

HOGARTH

suitability of axioms underlying mathematical systems or models of rational behavior. The nal arbiter of such systems is the intuition of educated people, which almost by denition cannot be stable. The axioms underlying rational choice, for example, have clearly evolved across time (e.g., from expected value to expected utility, to subjective expected utility, etc.).

Educating Intuition In my 2001 book, I put forth the idea that intuition can be explicitly educated. Whereas little explicit research has been carried out on this topic,11 I believe that the basic principles outlined in Hogarth (2001) are still (tentatively) valid. The argument for the educability of intuition is based on the notion that intuition is largely the result of learning. Thus, if experience is organized such that people learn the right lessons from their interactions with the world, intuition can be educated in the sense that judgments can become more accurate in specic domains. In many ways, then, educating intuition is similar to dening the conditions that allow task-specic expertise to emerge. First, I need to make a parenthetical comment. We can really only talk normatively about this topic for intuitions that can subsequently be seen to be correct or incorrect, for example, in medical diagnosis or some prediction task. As previously noted, there is abundant evidence in society that organizations educate the intuitions (i.e., cultural capital) of their members. But whether one considers the outcomes of such education positively or negatively depends on personal taste (i.e., ones own cultural capital). In Hogarth (2001), I suggested seven guidelines for educating intuition. Of these, I consider three to be particularly important. They are (a) to select and/or create your environments, (b) seek feedback, and (c) make scientic method intuitive. These three guidelines differ in how active the participant needs to be in the educational process. The rst suggestion depends entirely on the power of automatic learning to form intuitions. Thus if you are interested in developing intuitive expertise in a particular domain, it pays to seek and/or participate in environments that are kind in the sense of providing you with the appropriate experience (see also Shanteau, 1992). Indeed, this approach is the basis of apprenticeships in many professions. That is, by shadowing recognized masters (be they musicians, craftsmen, managers, or even professors) people can acquire a sense of which reactions are appropriate in which situations.
11 For ideas on educating intuition in the context of management education, see Sadler-Smith and Burke (2009) and Sadler-Smith and Shefy (2007). For applications in the area of medicine, see Abernathy and Hamm (1994).

Related to this point is the notion that environments can be explicitly designed to facilitate intuitive processes. Thus, for example, people can recognize anomalies in the paths of moving objects if these are seen in dynamic, visual form (e.g., lms) but give incorrect answers when questions about the same phenomena are in the form of word puzzles (McCloskey, 1983). Similarly, as shown by Gigerenzer and Hoffrage (1995), Bayesian updating is considerably more accurate if the problem format highlights data in the form of natural frequencies as opposed to the usual probabilistic/verbal format. More generally, these and similar examples are all witnesses to what is now known as choice architecture (Thaler & Sunstein, 2008), which recognizes the interaction between task structures and the natural or intuitive ways that people have of making judgments. What is not clear, however, is whether learning in appropriately structured tasks leads to expertise that can transcend task presentation formats. How much generalization occurs? What are the boundary conditions? The second important guideline is to seek feedback. This may seem like stating the obvious, but it should be clear that many learning structures are such that feedback is either missing or misleading and that judgments are often implicitly evaluated by assuming that no news is good news. In Hogarth (2006), I conducted a study of what feedback people seek and receive from decisions in their daily lives. Eleven university students and 24 business executives participated in an experience sampling study (Hurlburt, 1997) in which they completed brief questionnaires in response to prompts made to their cellular telephones at random moments over a period of 1 or 2 weeks. Respondents were required to describe the last decision they had taken (which clearly could have been quite trivial such as taking a coffee after lunch) and how they knew or would know that the decision was appropriate, that is, to describe possible feedback. There were two main results. First, the percentage of decisions where people received or expected to receive feedback was not high (60%), and yet respondents indicated high condence in the correctness of their actions.12 Second, in many cases respondents were incapable of stating what evidence they could see to validate their decisions. In other words, they did not always know what feedback would be appropriate. Whereas these daily decision-making activities might seem removed from the types of tasks usually considered in research on intuitive processes, I believe that the results are illuminating because everyday decisions involve a mix of intuitive and deliberative
12 The types of decisions described by students and executives were very different as was the time horizon when feedback would become available. In general, the executives had to wait longer than students to receive feedback on their decisions.

Downloaded By: [Consorci de Biblioteques Universitaries de Catalunya] At: 12:45 3 December 2010

348

THE CHALLENGE OF INTUITION

Downloaded By: [Consorci de Biblioteques Universitaries de Catalunya] At: 12:45 3 December 2010

processes. Moreover, the processes used in these everyday tasks are likely to be similar to those used in making more important decisions where there may be no relevant feedback from the past. The third important guideline is to make scientic method intuitive. Clearly, this is a goal that is already pursued by many educational programs. In science classes, for example, people are taught how to reason using control groups, randomization, and so on. And yet, what is not clear is how the acquisition of such principles generalizes to other types of inferential tasks that people meet in their lives (see, e.g., Nisbett, Fong, Lehman, & Cheng, 1987). Just how well can we be trained to reason intuitively based on solid scientic principles?13 This point also relates to that made in the preceding section, namely, the need to train people to use both intuition and analysis in making decisions. Decision Making in a Changing World Given that intuitive responses have been honed on past experiences, future-oriented intuitive actions can clearly be effective only to the extent that the future resembles the past. For many situations, this is clearly the case. However, when the future differs signicantly from the past, reliance on intuitive processes can prove dysfunctional. Slovic et al. (2002) provided an interesting example with respect to the habit of smoking. Imagine a young adult who has acquired a liking for smoking such that the decision to smoke a cigarette on any particular occasion becomes an intuitive decision, that is, an automatic response to a perceived need. Moreover, this process will be repeated happily many times over the coming years. However, what the smoker does not realize at the outset is that smoking in later years carries a high cost in terms of negative consequences for health. Moreover, the very act of smoking changes the situation in that, across time, the person acquires a physical addiction. The point of this example is that many decisions that have an intuitive or emotional rationale may actually be dysfunctional because they inevitably ignore the fact that conditions can change. Moreover, whereas we can understand this kind of situation when discussing situations such as smoking or even the effects of certain diets, we have greater difculty with other topics such as how spending money today affects the savings we will need to nance our distant future selves. These concerns all raise an important issue: How do we train peoples intuitions to handle situations with which they are not familiar? The answer, in short, is that we cannot. Yet it is not clear that people can handle a future, changing world
13 For some speculative suggestions concerning how one might use e-mail to enable such a process, see Hogarth (2001, pp. 271272).

through deliberative processes. As an example, consider the well-documented inaccuracies in economic forecasts and particularly those that preceded the nancial crash of 2008. Is there any way then of using our intuitive capacities to help in the difcult task of handling unknowable futures? Let me make one practical suggestion based on a metaphor (Hogarth, 2010). The metaphor is that decision making across the life-cycle is like traveling across a mineeld in a fog. The general goalas in lifeis to get across the eld in good shape, and assume that in crossing the eld there are various positive rewards that you can collect. But at the same time, your ability to see where you are going is restrictedrandomly by the density of the fog so that, whereas you might be able to see quite far in some cases, this is not going to be the general rule. The mines, too, are distributed randomly around the eld and vary in how much damage they can inict. In the foggy mineeld, the target cannot be preciseits just to get to the other side in the best possible condition. As to subtargets, they are going to vary considerably depending on the state of the fog when you make any particular commitment. How then should one act? Clearly, there is no precise answer to this problem and our intuitions may not be helpful since we have never experienced such situations before. Our sole resort is the human ability to imagine the possible consequences of different future worlds. Fortunately, this may not be as hard as it rst seems because there are many technologies today that can help in this process. Consider, for example, using simulation models to stretch our imaginations and provide experiences of possible, future worlds. As a specic example, imagine a simulation model with the structural characteristics of foggy mineelds. This can be used in at least two ways. One is to observe how humans play such games, what kinds of rules they learn to adopt, and how successful they are (see, e.g., Broadbent, FitzGerald, & Broadbent, 1986). The second is to develop hypotheses about which types of decision rules (or intuitions?) are likely to be successful in different environments and to test these ideas explicitly through simulated games. Following the lessons of decision making gleaned from static environments, rules are likely to be successful to the extent that there is some match between their characteristics and those of the environments in which they operate (Hogarth & Karelaia, 2007). The outcomes of such investigations should be an appreciation for what thinking strategies or intuitions are suitable for environments we have not previously experienced.

Concluding Comments When asked in 1924 why he wanted to climb Mount Everest, the British mountaineer George Mallory 349

HOGARTH

reputedly answered, Because its there. Much the same could be said of why we should study intuition. It is present in almost all of our decision-making activities. In thisas well as in previous work (Hogarth, 2001)I dened intuition broadly as involving learned responses that are not the outcomes of deliberate processes. In contrast to previous research on decision making, which was heavily inuenced by Simons (1955) concept of bounded rationality and thus more focused on deliberate processes, the explicit incorporation of automatic and nonconscious processes is illuminating in that it has greatly expanded the range of decision making phenomena studied (Weber & Johnson, 2009). This is both welcome and challenging. Among other outcomes, it has led to a more optimistic view of human cognitive abilities. It can also lead to a better understanding of when intuition or analysis yields better decisions although, as previously argued, this question is better framed by asking how both modes of thought can best be combined. In this article I identied different areas of research where, I believe, attention to the intuitive dimension of thought could be especially instructive. Two of these were related to the notion of intuition as cultural capital and dealt with the evolution of preferences and moral judgments. The third raised the important issue that people can be active in educating their own intuitionsby, for example, exposing themselves to kind environments where they can benet from appropriate tacit learning. The fourth area recognizes that, because intuitions are based on past experience, there is a need to train people how to make use of intuitive abilities in a changing world. To date, a broad view of intuitive processes has proved useful for linking different areas of psychological research in decision making (cf. Hodgkinson et al., 2008). However, there is a danger that denitions of intuition such as that used in the present article are too inclusive. If intuition covers almost everything then perhaps it is nothing? Indeed, one section of the article was devoted to considering different functions and varieties of intuition and welcomed work that is helping to dene novel ways of classifying intuitive processes. Going forward, I believe that the greatest challenge facing intuition researchers is to determine more precise ways of classifying different types of intuitive phenomena that can lead to more specic and illuminating questions. It is time to move beyond simple dual-process concepts of thinking.

providing insightful comments on an earlier version of this article: Andreas Gl ockner, Danny Kahneman, Josh Klayman, Darcia Narvaez, Eugene Sadler-Smith, Emre Soyer, and Cilia Witteman. Errors and omissions are, of course, all mine. Note Address correspondence to Robin M. Hogarth, Universitat Pompeu Fabra, Department of Economics & Business, Ramon Trias Fargas 25-27, 08005 Barcelona, Spain. E-mail: robin.hogarth@upf.edu

Downloaded By: [Consorci de Biblioteques Universitaries de Catalunya] At: 12:45 3 December 2010

References
Abernathy, C. M., & Hamm, R. M. (1994). Surgical scripts: Master surgeons think about 43 common surgical problems. Philadelphia, PA: Hanley & Belfus. Acker, F. (2008). New ndings on unconscious versus conscious thought in decision making: Additional empirical data and meta-analysis. Judgment and Decision Making, 3, 292303. Bargh, J. A., & Morsella, E. (2008). The unconscious mind. Perspectives on Psychological Science, 3(1), 7379. Bazerman, M. H. (2005). Judgment in managerial decision making (6th ed.). Hoboken, NJ: Wiley. Bilali, M., McLeod, P., & Gobet, F. (2010). The mechanism of the Einstellung (set) effect: A pervasive source of cognitive bias. Current Directions in Psychological Science, 19, 111115. Blattberg, R. C., & Hoch, S. J. (1990). Database models and managerial intuition: 50% model + 50% manager. Management Science, 36, 887899. Bowers, K. S., Regehr, G., Balthazard, C., & Parker, K. (1990). Intuition in the context of discovery. Cognitive Psychology, 22, 72110. Briggs, K. C., & Myers, I. B. (1976). MyersBriggs type indicator. Palo Alto, CA: Consulting Psychologists Press. Broadbent, D. E., FitzGerald, P., & Broadbent, M. H. P. (1986). Implicit and explicit knowledge in the control of complex systems. British Journal of Psychology, 77, 3350. Brunswik, E. (1952). The conceptual framework of psychology. Chicago, IL: The University of Chicago Press. Camerer, C. F. (1981). General conditions for the success of bootstrapping models. Organizational Behavior and Human Performance, 27, 411422. Chaiken, S., & Trope, Y. (Eds.). (1999). Dual-process theories in social psychology. New York, NY: Guilford. Chapman, L. J., & Chapman, J. P. (1967). Genesis of popular but erroneous psycho-diagnostic observations. Journal of Abnormal Psychology, 72, 193204. Chase, W. G., & Simon, H. A. (1973). Perception in chess. Cognitive Psychology, 4, 5581. Cohen, L. J. (1981). On the psychology of prediction: Whose is the fallacy? Cognition, 7, 385407. Damasio, A. R. (1994). Descartes error: Emotion, reason and the human brain. New York, NY: HarperCollins. Dane, E., & Pratt, M. G. (2007). Exploring intuition and its role in managerial decision making. Academy of Management Review, 32, 3354. Dane, E., & Pratt, M. G. (2009). Conceptualizing and measuring intuition: A review of recent trends. In G. P. Hodgkinson & J. K. Ford (Eds.), International review of industrial and organizational psychology (Vol. 24, pp. 140). Chichester, UK: Wiley & Sons.

Acknowledgments Support from the Spanish Ministerio de Ciencia y Innovac on (grant ECO2009-09834) is gratefully acknowledged. I am indebted to several colleagues for 350

THE CHALLENGE OF INTUITION Denrell, J. (2005). Why most people disapprove of me: Experience sampling in impression formation. Psychological Review, 112, 951978. De Vries, M., Holland, R., & Witteman, C. L. M. (2008a). Fitting decisions: Diffuse affect and intuitive versus deliberative decision strategies. Cognition and Emotion, 22, 931943. De Vries, M., Holland, R., & Witteman, C. L. M. (2008b). In the winning mood: Affect in the Iowa gambling task. Judgment and Decision Making, 3, 4250. Dijksterhuis, A., Bos, M. W., Nordgren, L. F., & van Baaren, R. B. (2006). On making the right choice: The deliberation-withoutattention effect. Science, 311, 10051007. Dijksterhuis, A., & Nordgren, L. F. (2006). A theory of unconscious thought. Perspectives on Psychological Science, 1, 95109. Dougherty, M. R. P., Gettys, C. F., & Ogden, E. E. (1999). MINERVADM: A memory process model for judgments of likelihood. Psychological Review, 106, 180209. Einhorn, H. J. (1986). Accepting error to make less error. Journal of Personality Assessment, 59, 387395. Einhorn, H. J., & Hogarth, R. M. (1978). Condence in judgment: Persistence of the illusion of validity. Psychological Review, 85, 395416. Elwin, E., Juslin, P., Olsson, H., & Enkvist, T. (2007). Constructivist coding: Learning from selective feedback. Psychological Science, 18, 105110. Epstein, S. (1994). Integration of the cognitive and the psychodynamic unconscious. American Psychologist, 49, 709724. Epstein, S., Pacini, R., DenesRaj, V., & Heier, H. (1996). Individual differences in intuitive-experiential and analytical-rational thinking styles. Journal of Personality and Social Psychology, 71, 390405. Ericsson, K. A., & Charness, N. (1994). Expert performance: Its structure and acquisition. American Psychologist, 49, 725 747. Fiedler, K. (2000). Beware of samples! A cognitive-ecological sampling approach to judgment biases. Psychological Review, 107, 659676. Fiedler, K., & Juslin, P. (2006). (Eds.). Information sampling and adaptive cognition. New York, NY: Cambridge University Press. Gigerenzer, G. (2007). Gut feelings: The intelligence of the unconscious. New York, NY: Viking Penguin. Gigerenzer, G., & Hoffrage, U. (1995). How to improve Bayesian reasoning without instructions: Frequency formats. Psychological Review, 102, 684704. Gigerenzer, G., Hoffrage, U., & Kleinb olting, H. (1991). Probabilistic mental models: A Brunswikian theory of condence. Psychological Review, 98, 506528. Gigerenzer, G., Todd, P. M., & the ABC Group. (1999). Simple heuristics that make us smart. New York, NY: Oxford University Press. Gilovich, T., Grifn, D., & Kahneman, D. (Eds.). (2002). Heuristics and biases: The psychology of intuitive judgment. New York, NY: Cambridge University Press. Gladwell, M. (2005). Blink: The power of thinking without thinking. London: Penguin. Gl ockner, A., & Betsch, T. (2008a). Modeling option and strategy choices with connectionist networks: Towards an integrative model of automatic and deliberate decision making. Judgment and Decision Making, 3, 215228. Gl ockner, A., & Betsch, T. (2008b). Multiple-reason decision making based on automatic processing. Journal of Experimental Psychology: Learning, Memory, and Cognition, 34, 1055 1075. Gl ockner, A., & Witteman, C. L. M. (2010a). Beyond dual-process models: A categorisation of processes underlying intuitive judgement and decision making. Thinking & Reasoning, 16(1), 125. Gl ockner, A., & Witteman, C. L. M. (Eds.). (2010b). Foundations for tracing intuition: Challenges and methods. Hove, UK: Psychology Press. Goldberg, L. R. (1970). Man versus model of man: A rationale, plus some evidence, for a method of improving on clinical inferences. Psychological Bulletin, 73, 422432. Goldstein, D. G., & Gigerenzer, G. (2002). Models of ecological rationality: The recognition heuristic. Psychological Review, 109, 7590. Groopman, J. (2007). How doctors think. New York, NY: Houghton Mifin. Grove, W. M., Zald, D. H., Lebow, B. S., Snitz, B. E., & Nelson, C. (2000). Clinical versus mechanical prediction: A meta-analysis. Psychological Assessment, 12, 1930. Haidt, J. (2001). The emotional dog and its rational tail: A social intuitionist approach to moral judgment. Psychological Review, 108, 814834. Haidt, J. (2010). Moral psychology must not be based on faith and hope: Commentary on Narvaez (2010). Perspectives on Psychological Science, 5, 182184. Haidt, J., & Joseph, C. (2004, Fall). Intuitive ethics: how innately prepared intuitions generate culturally variable virtues. Daedalus, pp. 5566. Hammond, K. R. (1996). Human judgment and social policy: Irreducible uncertainty, inevitable error, unavoidable injustice. New York, NY: Oxford University Press. Hammond, K. R., Hamm, R. M., Grassia, J., & Pearson, T. (1987). Direct comparison of the efcacy of intuitive and analytical cognition in expert judgment. IEEE Transactions on Systems, Man, and Cybernetics, SMC-17, 753770. Hasher, L., & Zacks, R. T. (1979). Automatic and effortful processes in memory. Journal of Experimental Psychology: General, 108, 356358. Hasher, L., & Zacks, R. T. (1984). Automatic processing of fundamental information: The case of frequency occurrence. American Psychologist, 39, 13721388. Hodgkinson, G. P., Langan-Fox, J., & Sadler-Smith, E. (2008). Intuition: A fundamental bridging construct in the behavioural sciences. British Journal of Psychology, 99, 127. Hogarth, R. M. (1975). Cognitive processes and the assessment of subjective probability distributions. Journal of the American Statistical Association, 70, 271289. Hogarth, R. M. (1981). Beyond discrete biases: Functional and dysfunctional aspects of judgmental heuristics. Psychological Bulletin, 90, 197217. Hogarth, R. M. (2001). Educating intuition. Chicago, IL: The University of Chicago Press. Hogarth, R. M. (2005). Deciding analytically or trusting your intuition? The advantages and disadvantages of analytic and intuitive thought. In T. Betsch & S. Haberstroh (Eds.), The routines of decision making (pp. 6782). Mahwah, NJ: Erlbaum. Hogarth, R. M. (2006). Is condence in decisions related to feedback? Evidence from random samples of real-world behavior. In K. Fiedler & P. Juslin (Eds.), Information sampling and adaptive cognition (pp. 456484). Cambridge, UK: Cambridge University Press. Hogarth, R. M. (2010). On subways and coconuts in foggy mine elds: An approach to studying future-choice decisions. In E. Michel-Kerjan & P. Slovic (Eds.), The irrational economist. Making decisions in a dangerous world (pp. 2129). New York, NY: Public Affairs Books. Hogarth, R. M. (in press). When simple is hard to accept. In P. M. Todd, G. Gigerenzer, & The ABC Research Group (Eds.), Ecological rationality: Intelligence in the world. Oxford, UK: Oxford University Press. Hogarth, R. M., & Karelaia, N. (2007). Heuristic and linear models of judgment: Matching rules and environments. Psychological Review, 114, 733758.

Downloaded By: [Consorci de Biblioteques Universitaries de Catalunya] At: 12:45 3 December 2010

351

HOGARTH Hogarth, R. M., Portell, M., & Cuxart, A. (2007). What risks do people perceive in everyday life? A perspective gained from the experience sampling method (ESM). Risk Analysis, 27, 1427 1439. Hogarth, R. M., Portell, M., Cuxart, A., & Kolev, G. I. (in press). Emotion and reason in everyday risk perception. Journal of Behavioral Decision Making. Hogarth, R. M., & Soyer, E. (2010). Experiencing sequentially simulated outcomes: A guide to improve statistical inference (Working paper). Universitat Pompeu Fabra, Barcelona, Spain. Hsee, C. K., & Rottenstreich, Y. (2004). Music, pandas, and muggers: On the affective psychology of value. Journal of Experimental Psychology: General, 133, 2330. Hurlburt, R. T. (1997). Randomly sampling thinking in the natural environment. Journal of Consulting and Clinical Psychology, 67, 941949. Isen, A. (1993). Positive affect and decision making, In M. Lewis & J. M. Haviland (Eds.), Handbook of emotions (pp. 261277). New York, NY: Guilford. Jung, C. (1926). Psychological types (H.G. Baynes, Trans.). London: Routledge & Keegan Paul. Kahneman, D., & Frederick, S. (2002). Representativeness revisited: Attribute substitution in intuitive judgment. In T. Gilovich, D. Grifn, & D. Kahneman (Eds.), Heuristics and biases: The psychology of intuitive judgment (pp. 4981). New York, NY: Cambridge University Press. Kahneman, D., & Klein, G. (2009). Conditions for intuitive expertise: A failure to disagree. American Psychologist, 64, 515526. Kahneman, D., Slovic, P., & Tversky, A, (Eds.). (1982). Judgment under uncertainty: Heuristics and biases. New York, NY: Cambridge University Press. Karelaia, N., & Hogarth, R. M. (2008). Determinants of linear judgment: A meta-analysis of lens model studies. Psychological Bulletin, 134, 404426. Khatri, N., & Ng, H. A. 2000. The role of intuition in strategic decision making. Human Relations, 53, 5786. Klayman, J., & Ha, Y.-W. (1987). Conrmation, disconrmation, and information in hypothesis testing. Psychological Review, 94, 211228. Klein, G. A. (1993). A recognition-primed decision (RPD) model of rapid decision making. In G. A. Klein, J. Orasanu, R. Calderwood, & C. E. Zsambok (Eds.), Decision making in action: Models and methods (pp. 138147). Norwood, NJ: Ablex. Klein, G. (1998). Sources of power: How people make decisions. Cambridge, MA: MIT Press. Klein, G. A., Orasanu, J., Calderwood, R., & Zsambok, C. E. (1993). Decision making in action: Models and methods. Norwood, NJ: Ablex. Kmett, C. A., Arkes, H. R., & Jones, S. K. (1999). The inuence of decision aids on high school students satisfaction with their college choice decision. Personality and Social Psychology Bulletin, 25, 12931301. Le Doux, J. E. (1996). The emotional brain. New York, NY: Simon and Schuster. Le Mens, G., & Denrell, J. (2010). Adaptive sampling and biased beliefs: A rational analysis. (Working paper). Universitat Pompeu Fabra, Barcelona, Spain. Lewicki, P., Hill, T., & Czyzewska, M. (1992). Nonconscious acquisition of information. American Psychologist, 47, 796801. Loewenstein, G. F., Weber, E. U., Hsee, C. K., & Welch, N. (2001). Risk as feelings. Psychological Bulletin, 127, 267286. Makridakis, S., Hogarth, R. M., & Gaba, A. (2009). Dance with chance: Making luck work for you. Oxford: Oneworld. McCloskey, M. (1983). Intuitive physics. Scientic American, 248, 122130. Meehl, P. E. (1954). Clinical vs. statistical prediction: A theoretical analysis and a review of the evidence. Minneapolis, MN: University of Minnesota Press. Myers, D. G. (2002). Intuition: Its powers and perils. New Haven, CT: Yale University Press. Narvaez, D. (2006). Integrative ethical education. In M. Killen & J. Smetana (Eds.), Handbook of moral development (pp. 703 733). Mahwah, NJ: Erlbaum. Narvaez, D. (2010). Moral complexity: The fatal attraction of truthiness and the importance of mature moral functioning. Perspectives on Psychological Science, 5, 163181. Newell, B. R., Wong, K. Y., Cheung, J. C. H., & Rakow, T. (2009). Think, blink, or sleep on it? The impact of modes of thought on complex decision making. Quarterly Journal of Experimental Psychology, 62, 707732. Nisbett, R. E., Fong, G. T., Lehman, D. R., & Cheng, P. W. (1987). Teaching reasoning. Science, 238, 625631. Payne, J. W., Samper, A., Bettman, J. R., & Luce, M. F. (2008). Boundary conditions on unconscious thought in complex decision making. Psychological Science, 19, 1118 1123. Peelen, M. V., Fei-Fei, L., & Kastner, S. (2010). Neural mechanisms of rapid natural scene categorization in human visual cortex. Nature, 460, 9498. Peters, J. T., Hammond, K. R., & Summers, D. A. (1974). A note on intuitive vs. analytic thinking. Organizational Behavior and Human Performance, 12, 125131. Plessner, T., Betsch, C., & Betsch, T. (Eds.). (2008). Intuition in judgment and decision making. Mahwah, NJ: Erlbaum. Reber, A. S. (1989). Implicit learning and tacit knowledge. Journal of Experimental Psychology: General, 118, 219235. Reber, A. S. (1993). Implicit learning and tacit knowledge: An essay on the cognitive unconscious. New York, NY: Oxford University Press. Sadler-Smith, E. (2008). Inside intuition. Abingdon, UK: Routledge. Sadler-Smith, E., & Burke, L. A. (2009). Fostering intuition in management education: Activities and resources. Journal of Management Education, 33, 239262. Sadler-Smith, E., & Shefy, E. (2007). Developing intuitive awareness in management education. Academy of Management Learning & Education, 6, 186205. Schein, E. H. (1985). Organizational culture and leadership: A dynamic view. San Francisco, CA: Jossey-Bass. Seligman, M. E. P. (1970). On the generality of laws of learning. Psychological Review, 77, 406418. Shanteau, J. (1992). Competence in experts: The role of task characteristics. Organizational Behavior and Human Decision Processes, 53, 252262. Simon, H. A. (1955). A behavioural model of rational choice. Quarterly Journal of Economics, 69, 99118. Simon, H. A., & Chase, W. G. (1973). Skill in chess. American Scientist, 61, 394403. Slovic, P., Finucane, M., Peters, E., & MacGregor, D. (2002). The affect heuristic. In T. Gilovich, D. Grifn, & D. Kahneman (Eds.), Intuitive judgment: Heuristics and biases (pp. 397420). New York, NY: Cambridge University Press. Stanovich, K. E., & West, R. F. (2000). Individual differences in reasoning: Implications for the rationality debate? Behavioral and Brain Sciences, 23, 645726. Stigler, G. J., & Becker, G. S. (1977). De gustibus non est disputandum. American Economic Review, 67(2), 7690. Thaler, R. H., & Sunstein, C. R. (2008). Nudge: Improving decisions about health, wealth, and happiness. New Haven, CT: Yale University Press. Tversky, A., & Kahneman, D. (1974). Judgment under uncertainty: Heuristics and biases. Science, 185, 11241131. Weber, E. U., & Johnson. E. J. (2009). Mindful judgment and decision making, Annual Review of Psychology, 60, 53 85. Wilson, T. D., Kraft, F., & Dunn, D. S. (1989). The disruptive effects of explaining attitudes: The moderating effect of knowledge

Downloaded By: [Consorci de Biblioteques Universitaries de Catalunya] At: 12:45 3 December 2010

352

THE CHALLENGE OF INTUITION about the attitude object. Journal of Experimental Social Psychology, 25, 379400. Wilson, T. D., & Schooler, J. W. (1991). Thinking too much: Introspection can reduce the quality of preferences and decisions. Journal of Personality and Social Psychology, 60, 181192. Zacks, R. T., & Hasher, L. (2002). Frequency processing: A twentyve year perspective. In P. Sedlmeier & T. Betsch (Eds.), Etc. frequency processing and cognition (pp. 2136). New York, NY: Oxford University Press. Zajonc, R. B. (1968). Attitudinal effects of mere exposure. Journal of Personality and Social Psychology Monograph Supplement, 9(2, Pt.2), 127. Zajonc, R. B. (1980). Feeling and thinking: Preferences need no inferences. American Psychologist, 35, 151175.

Downloaded By: [Consorci de Biblioteques Universitaries de Catalunya] At: 12:45 3 December 2010

353

S-ar putea să vă placă și