Attention, Information Processing and Choice in Incentive-Aligned Choice Experiments
Résumé
In incentive-alignment choice experiments, each decision is realized with some prob- ability prob < 1. Incentive alignment induces truth telling, i.e., respondents do not consciously lie, given the information they have processed. However, based on the psychological distance literature and the bounded rationality literature, we predict that prob < 1 is not necessarily enough to induce consumers to process information and choose as they would if choices were realized with certainty (prob = 1). In three eye tracking experiments, we vary the probability prob that choices will be realized, from 0 to 1, and study the impact on attention, information processing, and choice. Consistent with our hypotheses, we find that as prob is increased from 0 to 1, consumers: process the choice-relevant information more carefully and in a way that is more consistent with a compensatory decision process; become less novelty seeking; become more price sensitive. These findings have implications for the ecological validity of incentive-aligned preference measurement surveys. While it is not feasible to systematically use questions with high prob in the eld, we further predict and find that placing a higher-probability question (such as an external validity task) at the beginning rather than the end of a questionnaire has a desirable carryover effect on attention, information processing and choice throughout the questionnaire.