Comparing a multi-item Rasch scale for purchase intentions with preference elicitation using experimental auctions for consumer acceptance of biofortified food: An in-store study in Rwanda

Consumer acceptance of biofortified vitamin-A-rich products by urban populations in developing countries is an important preventive pathway for addressing vitamin A deficiencies by easing nutritional availability and access. The present study used data from a multi-variable in-store consumer study in Rwanda with bread and a snack product (mandazi) to develop and test a measure for stated purchase intentions based on Rasch scaling, and investigates the structural relationship between purchase intentions and multiple measures of potential drivers for purchase intentions. The study investigates the extent to which the purchase intent influence the decisions and monetary product preferences within an in-store Becker-DeGroot-Marschak (BDM) auction based on a product exchange to the biofortified product using the self-selected non-fortified product as reference. The key finding is that the nature of the relationship between the methods to elicit preferences is product-specific because of differences in the outcome of the BDM. The nature of the relationship also depends on whether the measurement uncertainty in the Rasch score for person locations for purchase intentions is considered. Including the measurement uncertainty reverses the extent to which the choice to bid and the bid amount within the BDM are driven by separate processes. Moreover, while actual liking helped predict purchase intentions, no evidence was found of such association for nutritional beliefs. Sensory attributes for the vitamin-A-biofortified products do not serve as key product features to promote (or detract from) purchase intentions. Overall, these findings indicate that auction-based methods and Rasch scales for purchase intentions are complementary measures in product research. • Non-hypothetical data were obtained from an in-store field auction. • Product-specific differences were found between purchase intentions and WTP. • Accounting for measurement uncertainty is key to the validity of purchase intentions. • Liking evaluation was predictive of purchase intentions. • WTP by BDM and Rasch scales are complementary measures in product research.