Cognitive Psychology Of Decision Biases Research Paper

1. Normative And Descriptive Analyses

The study of decision making is an interdisciplinary enterprise involving economics, political science, and psychology, as well as statistics and philosophy. One can distinguish two approaches to the analysis of decision making: normative and descriptive. Central to many of the social sciences is the normative approach, which assumes a rational decision maker with well-defined preferences. Every option is assumed to have a subjective ‘utility’ for the individual, a well-established position in his or her preference ordering (von Neumann and Morgenstern 1947). Because preferences are clear and stable they are expected to be invariant across normatively equivalent assessment methods (‘procedure invariance’), and across logically equivalent ways of describing the options (‘description invariance’). This conception, known as the rational theory of choice, is based primarily on a priori considerations and on normative assumptions rather than on empirical observation. Because the normative treatment specifies clear and compelling principles of rational behavior, it serves as a benchmark against which descriptive studies of decision making are compared.

The descriptive approach to individual decision making is based on empirical observation and experimental studies of choice behavior, which emphasize the role of information processing in people’s decisions. The empirical evidence is that people’s decisions often depart from the normative benchmark in systematic ways which can be viewed as decision biases. In essence, preferences appear to be formed, not merely revealed, during their elicitation, and their formation is often influenced by factors that ought not to matter from a normative perspective. The formation of preferences appears to depend on subtleties such as problem frame, method of elicitation, and the valuations and attitudes that these trigger. In what follows, selected empirical findings about individual decision making are presented and several cognitive psycho-logical principles that underlie the decision-making process, and lead to systematic bias, are described.

2. Diminishing Sensitivity, Risk Attitudes, And Frames

Decisions are often ‘risky’ in the sense that their outcomes are not known with certainty. Presented with a choice between a risky prospect that offers a 50 percent chance to win $200 (otherwise nothing) and an alternative of receiving $100 for sure, most people prefer the sure gain over the gamble, although the two prospects have the same expected value. (Expected value is the sum of possible outcomes weighted by their probability of occurrence.) Preference for a sure outcome over a risky prospect of equal expected value is called risk averse; indeed, people tend to be risk averse when choosing between prospects with positive outcomes. The tendency towards risk aversion can be explained by the notion of diminishing sensitivity, first formalized by Daniel Bernoulli in 1738. Just as the impact of a candle is greater when it is brought into a dark room than into a room that is well lit so, suggested Bernoulli, the utility resulting from a small increase in wealth will be inversely proportional to the amount of wealth already in one’s possession. It has since been assumed that people have a subjective utility function, and that preferences should be de-scribed using expected utility instead of expected value. According to expected utility, the worth of a gamble offering a 50 percent chance to win $200 (otherwise nothing) is 0.50 * u($200), where u is the person’s concave utility function. (A function is concave or convex if a line joining two points on the curve lies entirely below or above the curve, respectively.) It follows from a concave function that the subjective value attached to a gain of $100 is more than 50 percent of the value attached to a gain of $200, which entails preference for the sure $100 gain and, hence, risk aversion.

Consider now a choice between losses. When asked to choose between a prospect that offers a 50 percent chance to lose $200 (otherwise nothing) and the alternative of losing $100 for sure, most people prefer to take an even chance at losing $200 or nothing over a sure $100 loss. This is because diminishing sensitivity applies to negative as well as to positive outcomes: the impact of an initial $100 loss is greater than that of the next $100. This results in a convex function for losses and a preference for risky prospects over sure out-comes of equal expected value, called risk seeking. With the exception of prospects that involve very small probabilities, risk aversion is generally observed in choices involving gains, whereas risk seeking tends to hold in choices involving losses.

These risk attitudes form part of prospect theory (Kahneman and Tversky 1979), an influential descriptive theory of choice, which incorporates an S-shaped value function that is concave for gains and convex for losses. This function is also steeper for losses than for gains: Thus, a loss of $X is more aversive than a gain of $X is attractive, which is known as loss aversion. In addition, prospect theory is defined on gains and losses rather than total wealth, which captures the fact that people typically treat outcomes as departures from a current reference point, rather than in terms of final assets. These properties seem compelling and unobjectionable, yet their combination can lead to problematic consequences.

In one example (Tversky and Kahneman 1986), respondents are asked to assume themselves $300 richer and invited to choose between a sure gain of $100 or an equal chance to win $200 or nothing. Alternatively, they are asked to assume themselves $500 richer, and made to choose between a sure loss of $100 and an equal chance to lose $200 or nothing. In accord with the properties described above, most subjects choosing between gains are risk averse and prefer the certain $100 gain, whereas most subjects choosing between losses are risk seeking, preferring the risky prospect over the sure $100 loss. The two problems, however, are essentially identical: When the initial $300 or $500 wealth is added to the respective outcomes, both problems amount to a choice between $400 for sure vs. an even chance at $300 or $500. This notwithstanding, those choosing between what are framed as gains exhibit a preference for the former outcome, whereas those facing what are framed as losses prefer the latter. This is known as a framing effect. It occurs when alternative descriptions of what is essentially the same decision problem give rise to discrepant attitudes and to predictably different choices. Framing effects have been observed in dozens of studies involving economic, medical, and consumer decisions, among others.

3. Loss Aversion And The Status Quo Bias

A fundamental fact regarding people’s reaction to outcomes is loss aversion: The loss of utility associated with giving up a good is greater than the utility associated with obtaining it (Tversky and Kahneman 1991). An immediate implication of loss aversion is that people will not accept an even chance to win or lose $X, because the loss of $X is more aversive than the gain of $X is attractive. Indeed, people are generally willing to accept an even-chance prospect only when the gain is substantially greater than (about twice as large as) the loss. Loss aversion entails that the loss of utility associated with giving up a good that is in our possession is generally greater than the utility gain associated with obtaining that good. This yields ‘endowment effects ,’ wherein the mere possession of a good (thus viewing it as a potential loss) can lead to higher valuation of it than if it were not in one’s possession (Kahneman et al. 1990).

A closely related manifestation of loss aversion is a general reluctance to trade, which is illustrated in a study (Knetsch 1989) in which subjects were divided into two groups: Half of the subjects were given a decorated mug, and the others were given a large bar of Swiss chocolate. Later, each subject was shown the alternative gift, and offered the opportunity to trade their gift for the other. Because the initial allocation of gifts was arbitrary and transaction costs minimal, economic theory predicts that about half the subjects should exchange their gifts. On the other hand, if losses loom larger than gains, then most participants will be reluctant to give up the gift in their possession (a loss) in order to obtain the other (a gain). Indeed, only 10 percent of the participants chose to trade their gifts. This contrasts sharply with the 50 percent predicted by standard economic analysis in which the value of a good does not change when it becomes part of one’s endowment.

Loss aversion entails a strong tendency to maintain the status quo, because the disadvantages of departing from it loom larger than the advantages of its alternative (Samuelson and Zeckhauser 1988). A striking framing effect which relies on people’s tendency to maintain the status quo has been observed in the context of insurance decisions, when New Jersey and Pennsylvania both introduced the option of a limited right to sue, which entitles automobile drivers to lower insurance rates. The two states differed, however, in what they offered consumers as the default option. New Jersey motorists had to acquire the full right to sue (transaction costs were minimal: a sig-nature), whereas in Pennsylvania the full right to sue was the default. Presented with the choice, only about 20 percent of New Jersey drivers chose to acquire the full right to sue, while approximately 75 percent of Pennsylvania drivers chose to retain it. The difference in adoption rates due to the different frames had financial repercussions estimated at around $200 million dollars (Johnson et al. 1993).

Loss aversion promotes stability rather than change by inducing people to maintain their current position. Among other things, the reluctance to change induced by loss aversion can hinder the negotiated resolution of disputes. If each side to a dispute evaluates the opponent’s concessions as gains and its own concessions as losses, then agreement will be hard to reach because each side will perceive itself as relinquishing more than it stands to gain.

4. Compatibility And Reversals

An important cognitive principle that relates the characteristics of stimuli to the ways in which people code the information and produce a response is the principle of compatibility. This principle has long been recognized by students of perception and motor control. It states that when stimuli and responses are mentally represented, the weight of a stimulus at-tribute is enhanced to the extent that it is compatible with the required response. Alternative methods of eliciting preference can thus lead to differential weighings of attributes and, consequently, to different choices.

In line with compatibility, for example, setting the price of a gamble is likely to emphasize payoffs more than probabilities because both the price and the payoffs are in monetary units (e.g., dollars). Consequently, a gamble’s potential payoff will be weighted more heavily in a pricing task than in a choice between gambles. Consistent with this is the preference reversal phenomenon (Slovic and Lichtenstein 1983), wherein subjects choose a lottery that offers a greater chance to win over another which offers a higher payoff, but then price the latter higher than the former. This pattern has been observed in numerous experiments, including one involving professional gamblers in a Las Vegas casino (Lichtenstein and Slovic 1973), and another offering the equivalent of a month’s salary to respondents in the Peoples’ Republic of China (Kachelmeier and Shehata 1992).

For another example, consider having to choose one of two options or, alternatively, having to reject one of two options. The two tasks are logically equivalent: If people prefer the first they should reject the second, and vice versa. However, because it is natural to select options based on their strengths, and to reject options based on their weaknesses, compatibility suggests that the strengths of options will loom larger when choosing, whereas options’ weak-nesses will be weighted more heavily when rejecting.

In one study (Shafir 1993), subjects were presented with pairs of options, one of which-–the ‘enriched’ option—had more strengths and more weaknesses than the other, ‘impoverished’ option. Because options’ strengths are weighted more heavily when choosing than when rejecting, and their weaknesses are weighted more heavily when rejecting than when choosing, the enriched option was chosen and rejected relatively more often than the impoverished option. Contrary to procedure invariance, logically equivalent tasks give rise to predictably different choices.

5. Independent And Comparative Evaluation

Some procedures, such as pricing or rating, involve an independent evaluation of each option, whereas others, such as choice, are inherently comparative in nature. Independent vs. comparative procedures tend to highlight different aspects of options and can thus alter decision. As it turns out, people feel that some attributes (e.g., safety) are more important than others (e.g., cost). There is evidence that the attribute that is judged as more important looms larger in choice, where it can act as a tie-breaker, than in independent evaluation, where its comparative advantage is less prominent (Tversky et al. 1988). This can lead to systematic violations of invariance.

One study, for example, elicited people’s support for interventions addressing environmental problems (Kahneman and Ritov 1994). One intervention concerned human health; for example, free medical checkups for farm workers facing the risk of skin cancer. The other concerned environmental protection: the contribution to a fund to provide safe breeding areas for Australian mammal species nearly wiped out by hunters. One group of respondents were asked to choose which of the two interventions they would rather support; a second group were presented with these one at a time and asked to determine the largest amount they would be willing to pay for each intervention. When asked to evaluate each intervention separately, respondents, who were predictably more moved by the hunted animals’ plight than by mundane checkups, were willing to pay more, on average, for the safe breeding of the Australian mammals than for skin cancer checkups. However, when faced with a direct choice between these options, most subjects favored checkups for humans over safe breeding for animals. As expected, the issue viewed as more important (cancer in humans rather than breeding of animals) acquired greater prominence in choice, which involves direct comparison between issues, than in separate evaluation, where other considerations prove influential absent a direct comparison.

A related pattern emerges when attributes are difficult to gauge in isolation. Hsee (1996), for example, presented subjects with two second-hand music dictionaries, one with 20,000 entries and a slightly torn cover; the other with 10,000 entries and a cover that was like new. Subjects had little notion of how many entries to expect in a music dictionary. Under separate evaluation, they were willing to pay more for the dictionary with the new rather than the torn cover. When these were evaluated concurrently, however, most subjects obviously preferred the dictionary with twice as many entries, despite its inferior cover.

6. Preferences Constructed In Context

People do not always have clear and well-ordered preferences: Instead, they approach decisions as problems that need to be solved, and construct preferences that are influenced by the context of decision. Among other things, decision is thus biased by the ways in which the options are described and by the methods through which preference is elicited. Other aspects of decision that have been found to yield bias include the tendency to delay decision when choices are hard (Tversky and Shafir 1992), the regret anticipated when another option could prove superior (Bell 1982), the influence exerted by costs already incurred (Arkes and Blumer 1985), the effects of temporal separation on future decisions (Loewenstein and Elster 1992), the role of reasons and justification (Shafir et al. 1993, Tetlock 1992), and the occasional inability to predict future tastes or to remember past satisfaction (Kahneman 1994).

Decision biases have been documented in real settings and with experts; for example, in physicians’ decisions regarding patient treatment (McNeil et al. 1982, Redelmeier and Shafir 1995), academics’ retirement investment decisions (Benartzi and Thaler 1995), and taxi drivers’ allocation of working hours (Camerer et al. 1997). Because these biases emerge from underlying cognitive mechanisms, they form an inherent part of the ways in which we process information and make decisions.

Bibliography:

1. Arkes H R, Blumer C 1985 The psychology of sunk cost. Organizational Behavior and Human Performance 35: 129–40
2. Bell D E 1982 Regret in decision making under uncertainty. Operations Research 30: 961–81
3. Benartzi S, Thaler R 1995 Myopic loss aversion and the equity premium puzzle. Quarterly Journal of Economics 110(1): 73–92
4. Camerer C F 1995 Individual decision making. In: Kagel J H, Roth A E (eds.) Handbook of Experimental Economics. Princeton University Press, Princeton, NJ, pp. 587–703
5. Camerer C, Babcock L, Loewenstein G, Thaler R 1997 Labor supply of New York City cab drivers. One day at a time. Quarterly Journal of Economics 112(2): 407–42
6. Hsee C K 1996 The evaluability hypothesis: An explanation of preference reversals between joint and separate evaluations of alternatives. Organizational Behavior and Human Decision Processes 67: 247–57
7. Johnson E J, Hershey J, Meszaros J, Kunreuther H 1993 Framing, probability distortions, and insurance decisions. Journal of Risk and Uncertainty 7: 35–51
8. Kachelmeier S J, Shehata M 1992 Examining risk preferences under high monetary incentives: Experimental evidence from the People’s Republic of China. American Economic Review 82: 1120–41
9. Kahneman D 1994 New challenges to the rationality assumption. Journal of Institutional and Theoretical Economics 150(1): 18–36
10. Kahneman D, Knetsch J L, Thaler R 1990 Experimental tests of the endowment effect and the Coase theorem. Journal of Political Economy 98(6): 1325–48
11. Kahneman D, Ritov I 1994 Determinants of stated willingness to pay for public goods: A study in the headline method. Journal of Risk and Uncertainty 9: 5–38
12. Kahneman D, Tversky A 1979 Prospect theory: An analysis of decision under risk. Econometrica 47: 263–91
13. Knetsch J L 1989 The endowment effect and evidence of nonreversible indifference curves. American Economic Review 79: 1277–84
14. Lichtenstein S, Slovic P 1973 Response-induced reversals of preference in gambling: An extended replication in Las Vegas. Journal of Experimental Psychology 101: 16–20
15. Loewenstein G, Elster J (eds.) 1992 Choice Over Time. Russell Sage Foundation, New York
16. McNeil B J, Pauker S G, Sox H C, Tversky A 1982 On the elicitation of preferences for alternative therapies. New Eng-land Journal of Medicine 306: 1259–62
17. Redelmeier D, Shafir E 1995 Medical decision making in situations that offer multiple alternatives. Journal of the American Medical Association 273(4): 302–5
18. Samuelson W, Zeckhauser R 1988 Status quo bias in decision making. Journal of Risk and Uncertainty 1: 7–59
19. Shafir E 1993 Choosing versus rejecting: Why some options are both better and worse than others. Memory and Cognition 21(4): 546–56
20. Shafir E, Simonson I, Tversky A 1993 Reason-based choice. Cognition 49(2): 11–36
21. Slovic P, Lichtenstein S 1983 Preference reversals: A broader perspective. American Economic Review 73: 596–605
22. Tetlock P E 1992 The impact of accountability on judgment and choice: Toward a social contingency model. Advances in Experimental Social Psychology 25: 331–76
23. Tversky A, Kahneman D 1986 Rational choice and the framing of decisions. Journal of Business 59(4): 251–78
24. Tversky A, Kahneman D 1991 Loss aversion in riskless choice: A reference dependent model. Quarterly Journal of Economics 106: 1039–61
25. Tversky A, Sattath S, Slovic P 1988 Contingent weighting in judgment and choice. Psychological Review 95(3): 371–84
26. Tversky A, Shafir E 1992 Choice under conflict: The dynamics of deferred decision. Psychological Science 3(6): 358–61
27. von Neumann J, Morgenstern O 1947 Theory of Games and Economic Behavior, 2nd edn. Princeton University Press, Princeton, NJ