Appealing dish names to nudge diners to more sustainable food choices: a quasi-experimental study

Abstract

Background: Promoting plant-rich diets, i.e., diets with significantly reduced amounts of animal products, including vegan and vegetarian, is a promising strategy to help address the dual environmental and health crises that we currently face. Appealing dish names could boost interest in plant-rich dishes by attracting diners' attention to them. In this study, a systematic approach to naming plant-rich dishes with appealing descriptors was tested with a quasi-experimental design in four workplace, self-service, buffet-style cafeterias in Chicago, Sydney, São Paulo and Singapore.

Methods: Three different plant-rich dishes were tested at each site. Appealing names were generated systematically through a workshop and emphasized the dish ingredients, origin, flavor and/or the eating experience. Each test dish appeared once in a four-week menu cycle where menu options changed on a daily basis. The cycle was then repeated four times (six times in Chicago) with the total number of showings for each dish to be four (six in Chicago). The dish names alternated between basic and appealing across dish repetitions. For each dish, the food taken per plate was estimated by weighing the overall food taken and dividing it by the plate count in the cafeteria. Data was analysed as percentage change from baseline (i.e., the first showing of each dish that always had a basic name) with linear mixed effects analysis using the lme4 package in R.

Results: Overall, appealing dish names significantly increased the amount of food taken per plate by 43.9% relative to baseline compared to basic dish names (54.5% vs. 10.6% increase for appealing vs. basic names, respectively, p = .002). This increase corresponded to a 7% increase in actual grams of food taken per plate. Secondary analysis showed that the effect was site-specific to English-speaking countries only and that there was no substitution effect between plant-rich and meat dishes.

Conclusions: The study tested an approach to creating appealing dish names in a systematic way and indicates that, in some settings, appealing dish titles are a relatively easy, scalable, cost-effective strategy that the food services sector can adopt to shift food choices towards more plant-rich, sustainable ones.

Keywords: Climate change; Dish titles; Food service; Language; Plant-rich diets; Sustainable food choices.

Fig. 1

Schematic Representation of the Workshop

Fig. 2

Percentage change of the food amount taken per plate relative to baseline for the basic (n = 28) and appealing (n = 30) names of plant-rich dishes. The food amount taken was estimated by dividing the total amount of food taken for each dish by the total plate count from that lunch period. The food amount taken was then normalized to the first presentation of that dish (which always had a basic name)

Fig. 3

Percentage change of the food amount taken per plate relative to baseline for meat dishes when the corresponding plant-rich dish names were either basic (n = 22) or appealing (n = 24). The food amount taken was estimated by dividing the total amount of food taken for each dish by the total plate count from that lunch period. The food amount taken was then normalized to the first presentation of that dish. Meat dishes were always presented with basic names

Fig. 4

Percentage change of the food amount taken per plate relative to baseline for the basic and appealing names of plant-rich dishes by study location. Sample sizes: Chicago n = 18 (8 basic names, 10 appealing names), São Paulo n = 12 (6 basic names, 6 appealing names), Singapore n = 12 (6 basic names, 6 appealing names), Sydney n = 16 (8 basic names, 8 appealing names). The food amount taken was estimated by dividing the total amount of food taken for each dish by the total plate count from that lunch period. The food amount taken was then normalized to the first presentation of that dish (which always had a basic name)

Fig. 5

Percentage change of the food amount taken per plate relative to baseline for the basic and appealing names of plant-rich dishes by dish type. Sample sizes: Main n = 28 (n = 14 for basic names, n = 14 for appealing names), Salad n = 4 (n = 2 for basic names, n = 2 for appealing names), Side n = 8 (n = 4 for basic names, n = 4 for appealing names), Soup n = 18 (n = 8 for basic names, n = 10 for appealing names). The food amount taken was estimated by dividing the total amount of food taken for each dish by the total plate count from that lunch period. The food amount taken was then normalized to the first presentation of that dish (which always had a basic name). Statistical tests were omitted for salads due to low sample size