Experienced difficulty in sequential decision making.
Jason Riis · Deep Blue (University of Michigan) · 2003
Seven studies were conducted to examine the role of experiential factors in sequential decisions. People often have to make several decisions in short sequence, but the vast majority of academic work on decision making has examined decisions made in isolation. Several mechanisms are described by which deliberation about a prior decision could influence choice on a subsequent one. The studies demonstrated the operation of one class of mechanism---feelings that result from deliberating about a prior decision can influence preferences in a subsequent decision. Studies 1--3 demonstrated an order effect whereby more status quo selection was observed on a difficult decision among subjects who were first asked to make a preliminary, unrelated difficult decision. Studies 2 and 3 ruled out the possibility that the effect was due to a reduced tolerance for assuming responsibility or to a priming mechanism whereby the prior decision invoked particular thoughts that were relevant to the second decision. The effects are best explained by an account which appeals to the experienced difficulty invoked by the prior decision. This experience of difficulty decreases the decision maker's tolerance for uncertainty in subsequent decisions. Studies 4 and 5 extended the effect to risk aversion. Subjects who had to make a preliminary easy decision were less likely to choose a sure thing than were subjects who had to make a preliminary hard decision. The effect was significant in Study 5. Study 6 attempted to examine whether or not trait uncertainty would work in the same way as state uncertainty (which is an aspect of experienced difficulty). It did not, and trait uncertainty did not predict state uncertainty. Study 7 attempted a more direct manipulation of uncertainty by having one group of subjects give reasons for their choice. Consistent with previous studies, this did increase the subjects' subjective confidence (or certainty) but it did not decrease their preference for a status quo option. To the contrary, it increased that preference, but it is suggested that this manipulation incidentally manipulated accountability, perhaps to a greater degree than uncertainty, and this may have caused the effect.