Satisficing when buying information

Abstract In two studies expected value, expected utility, and satisficing approaches to one's decision to buy information were compared. The task for the first study was to judge the maximum worth of information by bidding to reduce an ambiguous prospect to a risky one when given an option between a risky prospect and an ambiguous one. A within-subjects design was used in which, for the risky prospect, the probability of gain and expected value were varied. The results were inconsistent with both expected value and expected utility models, but were consistent with a satisficing strategy. The second study was similar to the first but with context changes in the stimuli and a slightly different decision task. The results showed that the effects are robust across task and context. The main behavioral effect is that people appear more willing to buy information that may decrease an expected loss than information that may increase an expected gain. A decision justification principle is consistent with that effect.