実験社会科学 - 実験が切り開く21世紀の社会科学 -

 

Petko Kusev (City University London)

Memory-biased preferences: The influence of accessibility on risky decision-making and judgments

 

Abstract:
We studied precautionary decisions where participants decided whether or not to adopt a specified precaution with a known cost in the face of a described risk. We compared the risks taken for precautionary decisions with those taken for equivalent monetary gambles. Applying Tversky and Kahneman's (1992) prospect theory to these data we find that the weighting function required to model precautionary decisions exhibit different properties than those required to model choices between monetary gambles. This result indicates a failure of the descriptive invariance axiom of expected utility theory. For risky prospects with moderate and high probabilities overweighting of probability is observed - a finding not anticipated by prospect theory. We hypothesize that precautionary decisions differ from monetary gambles because the former cue accessible features in memory. Accordingly, we report evidence that the accessibility of risky prospects influences participants' preferences and decision weights. Our results highlight a need for theories which differentiate between decisions about monetary gambles and other types of decision-making under risk and uncertainty. For judgments, the experiments show that judged frequencies of sequentially encountered stimuli are affected by certain properties of the sequence configuration. We find (a) a first-run effect whereby people overestimate the frequency of a given category when that category is the first repeated category to occur in the sequence and (b) a dissociation between judgments and recall; respondents may judge one event more likely than the other and yet recall more instances of the latter. Judged frequency of categories of items is influenced by the first run - which may reflect the operation of a judgment heuristic. The distribution of recalled individual items does not correspond to the frequency estimates, indicating that participants do not make frequency judgments by sampling their memory for individual items. We propose a simple strategy whereby respondents use the first run as a cue to frequency that accounts for this observation and other context effects on memory and judgment.--

報告風景