Assessing the sensitivity of information distortion to four potential influences in studies of risky choice

The emergence of a leading alternative during the course of a decision is known to bias the evaluation of new information in a manner that favors that alternative. We report 3 studies that address the sensitivity of predecisional information distortion and its effects in hypothetical risky decisions with regard to 4 potential influences: choice domain, repeated choice, memory requirements, and intermediate progress questions. In Experiment 1 (N = 515), the magnitude of information distortion was similar in 5 choice domains (varied between participants) involving monetary gambles, song downloads, frequent-flyer miles, political decisions, or medical decisions. Information distortion mediated the relationship between our manipulation of initial preferences and participants' final choices, with the magnitude of the indirect effect being roughly similar across domains. These results replicate and extend previous findings. Additionally, distortion decreased significantly over 4 similar decision problems (within participants), but remained significant in the fourth problem. In Experiment 2 (N = 214), information distortion increased significantly when previously viewed information remained available, apparently because reiterating that information strengthened emerging preferences. In Experiment 3 (N = 223), the removal of intermediate progress questions that measure information distortion and emerging preferences did not significantly affect final choices, again replicating previous results. We conclude that predecisional information distortion is a relatively stable and robust phenomenon that deserves a prominent role in descriptive theories of choice.