Changes in the Culture of Consumption During COVID-19: a Decision-Tree Model

Abstract

This study aimed to understand changes in the culture of consumption in the selected consumer seg-ments regarding COVID-19 during the lockdown. The empirical research was conducted with a sample of 534 households. To predict changes in consumer behavior, the two-step clustering method was used. We also use the decision-tree model. We created four selected segments with remarkable consumption changes on the basis of their opinions concerning three points of view, namely, behavior changes, de-mographic data, and self-beliefs. Consequently, we classified subject pertinence (i.e., single respon-dents) to single clusters by decision trees using various statistical calculations to place the chosen attri-bute within the node of the tree structure. The factors most statistically significant for inclusion in the cluster were willingness to give up personal freedom, conservatism, and preferring functionality to appearance/design, with the last significant factor being meeting deadlines. We find indications that the more people purchase online, the less they decrease expenses and support charities, or/and the more people solve things remotely, the less they restrict taking loans. The results will help in identifying future models in retail conditions.