This study aims to compare different front-of-package label designs, using two schemes: (1) High In and (2) Nutrition Info with each scheme having (1) a version with colors (i.e., green, yellow, and/or red) indicating level of nutrient content and (2) a black-and-white version. Additionally the Nutrition Info scheme will have a version that includes the percent Daily Value in black and white. Labels will be compared against a no-label control and one another.
Assessing the Impact of Front-of-package Nutrient Labels on Consumer Understanding and Behavior
This study aims to compare different front-of-package label designs, using two schemes: (1) High In and (2) Nutrition Info. The labels will have variations, such as color, and will be compared to a no-label control and one another. There will be a total of 6 experimental conditions: 1) High In label with color; 2) High In label in black and white; 3) Nutrition Info label with color; 4) Nutrition Info in black and white; 5) Nutrition Info label in black and white with percent Daily Value listed; and 6) a no-label control.
Participants will view a series of food-and-beverage products labeled (or not) according to condition. The three primary outcomes are: (1) correct assessment of high nutrient contents in food-and-beverage products, (2) perceived healthfulness of packaged foods and beverages, and (3) selection of at least one item high in at least one nutrient of concern (saturated fat, sodium, added sugar) in a shopping task. Perceived message effectiveness will be examined as a secondary outcome.
All statistical analyses will use two-tailed tests with a significance level of <0.05. Per CONSORT Guidelines, all models comparing labels will be bivariate, regressing outcomes on indicators for labeling condition. For continuous outcomes, linear regression models will be used, and for dichotomous outcomes, risk ratios (or probability ratios) will be estimated using Poisson regression with a robust error variance.
Effect modification of the labels on primary outcomes will be explored for type 2 diabetes status and sociodemographic variables important for health equity using interaction terms and stratified models.
Additionally, differences in the primary outcomes between each of the 6 unique conditions will be explored.
