Don't shop hungry? A receipt-based field study of hunger, shopping purpose, and unplanned purchasing

Hunger is widely assumed to drive unplanned grocery purchasing, yet direct behavioral evidence remains limited. This study examines the relationship between hunger, shopping purpose, and age as predictors of unplanned purchasing using receipt-based behavioral data from 323 shoppers and 3819 individually coded items. Using item-level generalized linear mixed models with binomial outcomes and random intercepts for shopper, hunger showed no significant association with basket-level unplanned purchasing (OR = 1.019, p = .434). However, category-level analysis controlling for age and shopping purpose revealed a significant domain-specific hunger effect in Snacks and Confectionery (OR = 1.090, p = .001), surviving both Bonferroni and FDR corrections, consistent with visceral influences theory. Shopping purpose was the dominant predictor (η ² = .223), with browsing shoppers showing nearly nine times the odds of an unplanned purchase relative to routine shoppers (OR = 8.635). Age exerted a significant negative effect, moderated by a significant hunger × age interaction (p = .014) with simple slopes confirming that hunger predicted unplanned purchasing only among younger shoppers. Latent profile analysis identified three distinct shopper profiles. Findings challenge the basket-level hunger-impulse assumption and position shopping purpose as a dominant structural predictor of unplanned grocery purchasing in naturalistic receipt-based data.