Modelling drivers of consumer liking handling consumer and product effects


  • Cristina Davino Department of Economics and Statistics, University of Naples Federico II, Naples, Italy
  • Rosaria Romano Department of Political and Social Sciences, University of Calabria, Cosenza, Italy
  • Domenico Vistocco Department of Political Sciences, University of Naples Federico II, Naples, Italy



drivers of liking, individual differences, product effect, quantile regression


The aim of the present paper is to approach the analysis of the relationship between overall liking and specific likings for eleven types of white corn tortilla chips. The main objective is to estimate a model for predicting the overall liking that also considers the heterogeneity in consumer liking. A further objective is to evaluate the adequacy of a single model for the different products. The first objective is achieved by using quantile regression, providing an estimate of the dependency relationship between overall and specific likings with respect to predefined quantiles, each corresponding to a specific segment of consumers. The second objective is achieved by using a specific strategy aimed at finding specific models for each product or group of similar products. The results show that the overall liking mostly depends on one specific liking, and its impact varies significantly for different segments of consumers. Furthermore, three different models are identified for three groups of products that differ in the same most important driver of the global model.