Balancing Samples in Stated Preference Studies

Abstract: There is a growing interest in quantifying the degree of heterogeneity in stated preferences for health. A popular investigation into preference heterogeneity involves split-sample analysis to make comparisons across subgroups. However, subgroups may differ in many observed characteristics and not accounting for these other characteristics may bias comparisons if these are also associated with preferences.

Likewise, recruiting respondents to stated preference studies can be challenging and the study sample may not be representative of the population of interest (e.g., a patient population, the general population). Unrepresentative samples may be useful for indicating preferences but if they differ in characteristics that are associated with preferences, decision-makers may struggle to generalise from the results.
This study explores how weighting and balancing approaches from the causal inference literature can be used in stated preference research to improve the balance of samples. This seminar will cover methods such as propensity-score weighting and entropy-balancing, and examine how these approaches can affect results and the subsequent conclusions drawn using hypothetical (simulated) and empirical examples.

The seminar will finish with an exploration of when these approaches should be used and when they may not be useful, including a discussion of how balancing samples may address barriers to the transportability of preference results for health care decision making.

Speaker Details:
Caroline is a Research Economist in the Health Preference Assessment team at RTI Health Solutions. Her research focuses on stated preference research, particularly using discrete choice experiments (DCEs) to elicit preferences for benefit-risk trade-offs. She has methodological interests in both the qualitative and quantitative components of stated preference research. She was invited to co-lead an International Society for Pharmacoeconomics and Outcomes Research (ISPOR) Webinar: Why and How to Use Qualitative Research Methods in Conjunction with DCEs, and recently first-authored an ISPOR Special Interest Group’s report on preference heterogeneity in DCEs. Prior to joining RTI Health Solutions, Caroline completed her PhD and post-doc research at the University of Manchester. As a PhD student she won the Lee B. Lusted Prize at the Society for Medical Decision Making’s (SMDM) 36th Annual Conference. Caroline has given oral presentations of her research at international conferences and has peer-reviewed publications including systematic reviews, editorials, empirical applications, and pedagogical pieces in journals such as PharmacoEconomics, Medical Decision Making, Value in Health, and the Journal of Choice Modelling. She currently sits on the honorary editorial board for The Patient journal.

REGISTER HERE