A person’s risk preference may determine significant life outcomes (e.g., in finance or health). People are therefore routinely asked to report their risk preferences in various scientific and applied contexts, yet still little is known concerning the cognitive underpinnings of this judgment-formation process. We ran two studies (N = 250, and N = 150 in a retest) implementing the process-tracing method of aspect listing to investigate the information- integration processes underlying people’s self-reports by means of cognitive modeling (RQ1) and to examine people’s cognitive representations of their risk preferences (RQ2). Our analyses indicate that interindividual differ- ences in self-reported risk preferences can be modeled well based on the listed aspects’ properties of evidence, and substantially better than using sociodemographic variables as predictors. Specifically, to render self-reports people appear to integrate the strength of evidence of multiple aspects sam- pled from memory. These aspects—that is, people’s cognitive representation of their risk preferences—mostly referred to the magnitudes of outcomes, and in line with a risk–return perspective, often explicitly referred to trade-offs between positive and negative outcomes. Crucially, within participants the strength of evidence of the listed aspects remained highly stable across the two studies (RQ3), and changes therein were closely related to changes in self-reported risk preference (RQ4). In sum, our findings provide cognitive insights concerning how people render self-reports of their risk preferences, suggest an explanation for the well-documented temporal stability thereof, and thus corroborate the internal validity of this measurement approach.
Through the window of my mind: Mapping information integration and the cognitive representations underlying self-reported risk preference.