Pre-Decisional Distortion of Risk Information Seen in Icon Arrays

In three studies, we tested whether icon arrays—which are a popular method of presenting risk information—can reduce pre-decisional information distortion that arises when early emerging preferences bias the evaluation of subsequently shown information. In Study 1, using traditional measures of information distortion, we found that risk-of-death information about two potential treatment options that was presented via icon arrays was distorted in favor of participants’ leading alternative. The magnitude of distortion was similar to the level of distortion for other treatment information in the treatment scenario. Study 2 directly tested whether the presence vs absence of icon arrays when presenting risk information had any impact on levels of information distortion, this time using a dependent measure that targeted people’s intuitive perceptions of risk. We found that the extent to which a 6% risk of death seemed riskier than a 3% risk-of-death was greater when the former risk was from a treatment option that was relatively undesired. This distortion was not significantly reduced by the presence of icon arrays. We replicated this pattern of results in a third study. These findings highlight the need for developing new tools and methods for presenting risk/likelihood information that can protect against the influence of pre-decisional information distortion.