Evaluation of two social norms nudge interventions to promote healthier food choices in a Canadian grocery store

Abstract

The objective of this study was to determine the impact of two nudge interventions on customers’ produce purchases at a rural Canadian grocery store. A pre- and post-intervention observational study design was used. Sales data were gathered before and after the staggered implementation of two nudge-based interventions to encourage produce purchases: grocery cart dividers to encourage shoppers to fill one-third of their cart with produce and grocery cart plaques with information about how many fruits and vegetables were typically purchased in the store. The proportion of total sales accounted for by produce was compared between baseline and implementation of the first intervention (Phase 1), between implementation of the first intervention and the addition of the second intervention (Phase 2), and between baseline and post-implementation of both interventions together. There was a 5% relative increase (0.5% absolute increase) in produce spending between baseline and post-implementation of both interventions (10.3% to 10.8%, p<0.001, 95% CI 0.2%, 0.7%). Intervention phase-specific produce spending showed no significant change in the percentage of produce spending from baseline to Phase 1 of the intervention, and an 8% relative increase (0.8% absolute increase) in the percentage of produce spending from Phase 1 to Phase 2 of the intervention (10.3% to 11.1%, p<0.001, 95% CI 0.5, 1.1%). Simple, low-cost nudge interventions were effective at increasing the proportion of total grocery spend on produce. This study also demonstrated that partnerships with local businesses can promote healthier food choices in rural communities in Canada.

Using social norm nudges in supermarket shopping trolleys to increase fruit and vegetable purchases

“…The higher mean sales of F&Vs by $9.10 (33.5% relative increase) per shopper in the social norm nudge intervention group exceeds evidence from the literature (Niculescu et al, 2016;Payne et al, 2015;Suleman et al, 2022). One reason for the higher spending on F&V sales compared to the literature could be that the current research used individual shopper receipts and not aggregated store sales data that may underestimate the increased purchases at an individual level.…”

“…However, this study did not report spending data or fruit purchases and shoppers in the control group were assessed on different data collection days than the intervention, potentially introducing a confounder since spending patterns can vary by day and time. Suleman et al (2022) also investigated the impact of two nudge interventions on shopper F&V purchases. Grocery carts were divided to encourage shoppers to fill one-third of their carts with F&Vs and grocery cart placards with information about how many F&Vs were typically purchased in the store were added to the trolley.…”

“…The mean total spend was $144.29 and $27.67 was spent on F&Vs. The calculation of the mean difference between groups was estimated from previous studies (Huitink et al, 2020;Niculescu et al, 2016;Payne et al, 2015;Suleman et al, 2022;Wansink et al, 2017). To produce a p-value of less than 0.05 at 95% power for a continuous endpoint in two independent samples of F&V spending (mean = $30 vs. $25, SD = $10), a sample size of 100 per group (Intervention and Control) was needed.…”