OBJECTIVES: A logical inconsistency occurs when an individual’s current preference logically contradicts a previous preference. These inconsistencies have been well documented in the literature, but this is the first paper of my knowledge examine logical inconsistencies within discrete choice experiments (DCEs) of the EQ-5D.

METHODS: I use responses from a nationally representative sample of 3,234 individuals from the US general population. As part of the survey, individuals were asked to complete a set of 30 paired comparisons involving the EQ-5D-5L. By comparing all individual responses, I determine whether or not an individual remained consistent in their preferences. I then examine the association between response consistency and individual characteristics including health status, household income, and educational attainment.

RESULTS: 52% of respondents suffered at least one logical inconsistency over the course of the experiment. Controlling for demographic variables, multivariate regressions find significant associations between logical inconsistencies and education, income, and problems with self-care. The findings show that on average, a one level increase in problems with self-care are associated with a 26% increase in the number of logical inconsistencies.

CONCLUSIONS: Certain sub-populations are more likely to provide inconsistent responses. Future research is needed to identify the source of these inconsistencies (e.g., learning, indifference, cognition). The advent of innovations in DCEs may be needed to increase the chances that individuals from lower socioeconomic status understand the preference tasks and are able provide accurate responses.