Food choice mimicry on a large university campus

Abstract

Social influence is a strong determinant of food consumption, which in turn influences the environment and health. Purchasing mimicry, a phenomenon where a person copies another person's purchases, has been identified as the key governing mechanism. Although consistent observations have been made on the role of purchasing mimicry in driving similarities in food consumption, much less is known about the precise prevalence, the affected subpopulations, and the food types most strongly associated with mimicry effects. Here, we study social influence on food choice through carefully designed causal analyses, leveraging the sequential nature of shop queues on a large university campus. In particular, we consider a large number of adjacent purchases where a focal user immediately follows another user ("partner") in the checkout queue and both make a purchase. Across food additions purchased during lunchtime together with a meal, we find that the focal user is significantly more likely to purchase the food item when the partner buys the item, vs. when the partner does not, increasing the purchasing probability by 14% in absolute terms, or by 83% in relative terms. The effect is observed across all food types, but largest for condiments. Furthermore, purchasing mimicry is present across age, gender, and status subpopulations, but strongest for students and the youngest. We elucidate the behavioral mechanism of purchasing mimicry, and derive direct implications for interventions improving dietary behaviors on campus, such as facilitating preordering to reduce detrimental interactions.

Keywords: campus; food choice; health; social influence; sustainability.

Fig. 1.

Study design. We identify dyads where two individuals make purchases within five minutes of each other, with no one in between, adjacent in the purchasing queue. The first person to make the transaction in the queue is referred to as the partner, and the second person as the focal person. We are interested in identifying the impact that the purchasing behavior of the partner has on the focal person e.g. purchasing a dessert as in the illustration. To that end, the dyads are matched, such that the dyads are comparable (i.e. they occur in the same shop, time of day, same partner identity, same availability, and popularity of the dessert), but, in treated dyads, the partner purchases a dessert, whereas in the control dyads, the partner does not purchase it. Our study then contrasts the focal person’s probability of purchasing the dessert, given that the partner purchased (treated) or not (control).

Fig. 2.

Purchasing mimicry across times of the day and the food items. In a) separately for lunch, breakfast, and afternoon or evening snack, the estimated RD (on the x-axis), for the different food item additions (on the y-axis). Risk difference estimates are marked with “o” and presented above (on the left: lunch where the anchor is the meal, on the right: breakfast and afternoon or evening snack where the anchor is a beverage). The randomized baseline is maked with “|” and presented below. In b) the estimated risk difference (on the x-axis), for the anchor type itself, as opposed to the food item addition (anchor type on the y-axis: type of meal, vegetarian vs. not, and type of beverage, coffee vs. tea). The error bars mark 95% bootstrapped CI.

Fig. 3.

Effect by the estimated status on campus. The estimated RD across the matched pairs of dyads (on the x-axis), depending on the individuals’ estimated status (on the y-axis). The error bars mark 95% bootstrapped CI. Risk difference estimates are marked with “o” and presented above, while the randomized baseline is marked with “|” and presented below. In a), for partner’s status, in b), for focal person’s status, in c) for the four combinations of the focal–partner status.

Fig. 4.

DAGs encode the assumptions about the causal relationship between variables. $X_a$ and $X_b$ are partner’s and focal person’s eating profile respectively; $S_{a,b}$ is the social tie strength; $Y_a(t)$ and $Y_b(t)$ are partner’s and focal person’s sets of purchased items at time t respectively, and $P(t)$ are common environmental factors at time t. Arrow from partner’s to focal person’s sets of purchased items marks the causal path of purchasing mimicry. In a), the assumed DAG. In b), c), and d), the variations of the assumed causal relationships where the Assumption 1 is violated such that the traits of the individuals can influence the observed purchasing behavior through factors not related to friendship strength $S_{a,b}$.