Influencing the nutritional quality of grocery purchases: A randomized trial to evaluate the impact of a social norm-based behavioral intervention with and without a loss-framed financial incentive

Abstract

Poor diet quality increases risks for non-communicable diseases. Interventions that leverage insights from psychology and economic theory are hypothesized to improve diet quality. This study tested the incremental effectiveness of these approaches using a fully operational online grocery store. We conducted a 3-arm crossover trial involving actual purchases with 187 primary grocery shoppers for households randomly exposed to: (1) Arm 1: control store, (2) Arm 2: store with a cost-free social norm-based behavioral intervention, built upon using Nutri-Score (NS), an evidence-based Front-of-Pack (FOP) interpretive label, and (3) Arm 3: Arm 2 plus a loss-framed financial incentive of SGD 5. Nutritional quality measures, including the average quality of the shopping basket based on the validated NS algorithm (primary), were used to assess intervention effectiveness using a linear mixed-effect model with repeated measures. The social norm-based behavioral intervention led to a statistically significant 4.62-point [95% CI: 3.73, 5.52] increase in the weighted average NS of the shopping basket, relative to Control. Adding the 5 SGD incentive increased effectiveness by an additional 2.49-point [CI: 1.61, 3.37]. These changes are equivalent to improving diet quality of the shopping basket, relative to Control, from NS grade low C to grade high C for Arm 2 and to Grade B for Arm 3. These findings suggest that interventions leveraging insights from behavioral science have the potential to improve nutritional quality at little to no additional cost and should be considered for adoption.