Combatting Truth-Telling Disincentives in Stated-Preference and Value-Set Surveys

Speaker(s)

Discussion Leader: Juan Marcos Gonzalez, PhD, Preference Evaluation Research Group, Duke Clinical Research Institute, Durham, NC, USA
Discussants: Reed Johnson, PhD, Preference Evaluation Research (PrefER) Group, Department of Population Health Sciences, Durham, NC, USA; David J Gebben, PhD, Center for Devices and Radiological Health, U.S. Food and Drug Administration, Silver Spring, MD, USA

PURPOSE:

This workshop will provide participants with tools to identify causes of uninformative or misleading responses in stated-preference and value-set surveys and guides participants through alternative strategies to encourage truthful responses. Participants become familiar with relevant findings on incentive-compatible study designs from market research and environmental economics and evaluate how to apply these findings in health applications.

DESCRIPTION:

When asked the question “What would you do if you had to make this choice?” a natural response could be: “This researcher is asking me to think hard about a speculative, emotionally uncomfortable situation with no reward for my effort.” This problem generally is known as “incentive incompatibility” in survey research. At best, noisy data from incentive-incompatible surveys produce imprecise parameter estimates. At worst, data from such surveys can be systematically biased from satisficing response strategies that economize on time and effort required to complete the surveys.

Workshop presenters will take 15 minutes each to discuss important considerations when designing incentive-compatible preference studies with methods such as “Bayesian truth serum” and game-theoretic “cheap talk.” Dr. Gebben will provide an overview of research on truth-telling in survey research, including the extensive research on improving veracity in nonmarket-valuation studies. Dr. Johnson will discuss implications for designing choice-experiment and threshold-method surveys for benefit-risk, stated-adherence, and value-set studies. Dr. Gonzalez will share results from surveys on Olympic-athlete doping decisions, preferences for novel weight-loss devices, and stated adherence to cystic-fibrosis antibiotic treatment.

Each audience member will respond to two versions of a survey to simulate the effects of incentive-incompatible and incentive-compatible survey designs. Panel members will lead a discussion comparing the results of the two designs. (15 minutes)

This workshop will be of considerable value to industry and regulatory users of preference data, as well as both less-experienced and more-experienced researchers.

Code

245

Topic

Patient-Centered Research