Decision Making In Prisoner’s Dilemma

5.2 Availability

Availability of instances or scenarios in memory, and constructive mental simulation serves for assessing frequency or probability of events. One may for example asses the risk of suffering a stroke by recalling instances of strokes among people that he knows or that he heard of, or by judging the plausibility of having a stroke using constructive mental simulation/imagination (some people may, then, fear suffering a stroke to an unreasonable degree and other people never think of it and judge the subjective probabilities accordingly, due to their different experiences or imaginations, although their actual probabilities can be almost identical).

5.21 Recall: differential availability

The problem with availability heuristic is that some information is more available – more readily retrievable from memory – than other information (for a comprehensive account see for example Kahneman & Frederick, 2002). Availability does not depend only on:

(a) the frequency of encountered instances (this frequency is of course not the same as the objective incidence, for example Kunreuther et al. (1978) found that people purchase flood insurance when they have personal experience with floods, while the knowledge of objective risks (and insurance rates) was imperfect);
but also on:
(b) salience or familiarity of information (Tversky & Kahneman, 1973/1982). That means that recall and consequently judgment of frequency is biased by dramatic or familiar instances, and does not reflect actual occurrence of all relevant instances. This non-objective and salience-biased recall can lead to stereotyping (Taylor, 1982), diagnostic failures, or overconfidence. Availability can be in various ways influenced by “feelings” and automatic unconscious processes (see for example Fazio, 2001; Bechara et al., 1997), but we are not going to concentrate on this issue in the present study.

5.22 Construction: prediction and attribution based on scenarios

When people want to predict future development they sometimes rely on scenarios – they model the possible development in mind, they construct a scenario and do not care for base rates or systematic checking of alternative developments. That is because some events are perceived as unique, or because decision makers do not have the base rate information at disposal or they simply ignore it. The perceived plausibility of a scenario (a plausible scenario seems “reasonable”, “logical”, “coherent”) or the difficulty of producing it serves as a clue to the likelihood of the occurrence of the event (Tversky & Kahneman, 1973/1982; Kahneman & Tversky, 1979/1982). And highly dramatic, salient, or “logical” scenarios are more available than is in accord with real incidence of cases in question (for example Singer, 1971/1982 show that property crimes committed by addicts are highly overestimated; Slovic et al., 1980/1982 shows that people relatively overestimate frequency of salient and dramatic causes of death such as tornadoes, flood, fire, homicide, or car accidents, and relatively underestimate causes such as diabetes, stroke, asthma, or stomach cancer – although there were several exceptions to the “availability” rule of assessment). Egocentrically driven and self-serving scenarios are also more available (Ross & Sicoly, 1979/1980; Ross & Anderson, 1982) – people attribute success to themselves and blame circumstances or other people for failure, they also succumb to the fundamental attribution error (overestimating personal factors in controlling behavior). “Theory-based” scenarios distort judgment as well (Jennings et al., 1982; Chapman & Chapman, 1971; Oskamp, 1965).

Note that modern neuroscience brings evidence of neurological and functional connection between recall and construction/simulation processes. Memories are not exact replicas of past events, instead, they offer material for creative investigation of future possible decisions, solutions and behavior. Recall and constructive processes are sometimes jointly termed “constructive memory” (see Dudai & Carruthers, 2005; Miller, 2007, Rossi et al., 2009). Differential, salient-biased, theory-based, or self-serving recall/simulation can build upon this neurological and cognitive connection of memory and construction/simulation.
Kuhlman et al., 1992 studied expectations of “cooperative”, “competitive” and “individualist” players (who were classified according to their actual decisions in the Prisoner’s Dilemma game) and discovered that players “projected” their own motives onto their opponents: cooperators expected their opponents to try to maximize joint pay-offs, competitors expected their opponents to try to maximize their own relative pay-off, while individualists expected their opponents to try to maximize their own absolute pay-off.

5.3 Anchoring

Anchoring means basing one’s estimates on some initially available value (e. g. a number). In one study (Tversky & Kahneman, 1974) subjects were given arbitrary numbers that markedly affected their estimates (in this example it was the number of African countries in the United Nations). Kahneman & Tversky (1974) also designed an interesting experiment of quick (within five seconds) numerical estimation of the product of the following multiplications (each equation was administered to one of two separate groups of students):

(A)

8 x 7 x 6 x 5 x 4 x 3 x 2 x 1

(B)

1 x 2 x 3 x 4 x 5 x 6 x 7 x 8

The estimated products were 2250 (A) and 512 (B). The median estimates of the two different groups were clearly anchored by the first few steps of the multiplication. For another example see Slovic et al., 1980/1982, pp. 481-482.
I believe anchoring can be viewed as a general tendency of human mind. It may be associated with such phenomena as priming (one’s reaction/response is subconsciously determined by prior stimuli) or therapeutic “orienting toward” (esp. in Ericksonian therapy, see for example Zeig & Geary, 2000, pp. xiv-xix; orienting toward means that a therapist inconspicuously, often rather vaguely, and/or metaphorically sets goals for a client that are to be reached in the future). Anchoring relies on more or less uncritical, quick, and subconscious acceptance of available anchors (numbers, perceptual stimuli, values, etc.).