Description-experience gap studies

Katherine Thompson

This project seeks to better understand the ways that everyday people use the probability of rare hazard events (like hurricanes, earthquakes, or volcanic eruptions), and how those uses of probability differ based on the source of the probability information. A better understanding of how people use such information under different circumstances will allow scientists and public officials to present hazard information to the public in ways that will more effectively promote preparedness behavior in cases where it is recommended, but also reduce overly high levels of concern where it is not warranted. The results from this project will allow scientists and emergency managers to improve the public’s understanding of the hazards in their environment, and allow public officials to create public policy that better synergizes with the ways that people use and understand probability, to maximize use and understanding of hazard information. Through its influence on decision architecture, this project could also affect the public’s behavior with respect to hazards. The specific objectives of the project are to better understand the Description-experience Gap, its boundary conditions, and the theory that (supposedly) underlies it, in both gains and losses and in novel or complex contexts; and create the structure of a virtual simulation game for testing the effect of real and virtual experience on homeowners’ behavior in reaction to debris-flow warnings.

Major Findings

• At least for choices that involve one certain option, it seems that a preference for the described-certain option strongly drives choices. To a lesser extent, avoidance of the described-risky option also drives choices.
• With sufficient sampling, it appears that choices between the two experienced decks are at or near chance. Process data shows no significant preference for either over the other, although there is a trend toward positive feelings toward E2, the certain experienced deck. Low rate of Ps noticing that the E decks matched the D decks is promising for planned fMRI studies, but as with previous results, it appears that more must be done to cloak the within-subject design from Ps. Data collection for the repeated-choice version of this study is ongoing at the moment.
• As of Month 4 of this project, our primary finding is a lack of support for underweighting theory in explaining the DE Gap. We see this in the DE Gap pilots: we tested the task design and replicated prior results, but novel comparisons in the within-subject design suggested explanations other than under-weighting (primarily uncertainty aversion), and process data further supported this conclusion.
• Further pilots indicated that the DE Gap may not be as robust under complex conditions as expected, and testing DE Gap effects with continuous outcomes found no support for under- and over-weighting per the Prospect Theory weighting function. Research is currently underway to better explain these unexpected results.

Broader impacts

• The results from this project will allow scientists and emergency managers to improve the public’s understanding of the hazards in their environment, and allow public officials to create public policy that better synergizes with the ways that people use and understand probability, to maximize use and understanding of hazard information.
• Through its influence on decision architecture, this project could also affect the public’s behavior with respect to hazards.

CRED2 Award (2010-2015): Funding was provided under the cooperative agreement NSF SES-0951516 awarded to the Center for Research on Environmental Decisions.