Dealing with ignored attributes through an inferred approach in wine choice experiments

A growing number of studies have sought to understand the effects that rules and heuristics might have on discrete choice models’ outputs, especially the impacts on willingness to pay measures. This paper presents a methodological approach to explore all possible strategies that respondents might use when making a choice of wine. Our results show that accounting for attribute non-attendance, that occurs when individuals ignore some product/service attributes in the choice process, improves the goodness-of-fit of models and their predictive capability. The inferred attendance approach developed in this article discloses that a minority of respondents considers all attributes and there is a multitude of attribute attendance possibilities that the researcher should search for. The use of heuristics by wine consumers highlights one implication of this research. Wineries should develop more appropriate communication strategies for consumers, especially with regard to the information provided on the labels. Additionally, it provides evidence that ignoring attribute non-attendance leads to overestimates of welfare measures.