Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
This paper explores the role of unobserved heterogeneity using different approaches: mixed, scaled and the generalized mixed logit model, and the latent class model, applied to wine choice. Results show that preference heterogeneity appears to be more relevant than scale heterogeneity, and the model accounting for both performs best. This finding suggests that wine choice is complex, demanding a higher cognitive burden, which may lead to the use of simplification strategies when choosing. For the purpose of wine valuation, the latent class model gives useful insights for market segmentation.