Risky-Choice Framing Effects Persist When Option Descriptions are Matched and Complete: A Replication and Extension of DeKay and Dou (2024)

People’s preferences in choices involving money, health, and other matters can often be steered toward safer or riskier options by framing the options as potential gains or potential losses. This risky-choice framing effect is well established, but its direction and magnitude depend on whether the options are described completely or incompletely and whether the descriptions are matched or mismatched in the gain and loss frames. In a recent experiment involving 18 combinations of option descriptions (nine in each frame), DeKay and Dou (2024) found that the valence or gist of the descriptions accounted for substantial variation in risk preferences, favoring categorical theories of framing over more mathematical theories. However, a residual framing effect and a framing effect in choices between completely described options contradicted the categorical theories and previous results. The framing effect for completely described options was especially surprising, but it was based on data from only 97 participants. The experiment reported here replicated all these effects in a large sample (N = 1,697) well-matched to U.S. Census data and with many more participants (477) making choices between completely described options. Risky-choice framing effects do stem partly from mismatched option descriptions in gains and losses, but a framing effect persists even when descriptions are matched and complete. The latter results expand the relevance of the effect to situations where it was thought to be absent. A new rescaled version of the explicated valence account explains these effects in the replication and in DeKay and Dou’s online and student samples.