The Impact of Nutrition Information and the Updated Nutrition Facts Label on Nutrition Literacy and Food Choices

Before completing a series of “food ranking tasks”, participants are randomly assigned to one of four experimental conditions in a 2 x 2 factorial design:
- Treatment 1: Nutrition Information
Participants either receive no nutrition information or are provided with general guidance on healthy eating at the beginning of the study.
- Treatment 2: Nutrition Facts Label
During the ranking tasks, participants are able to view either the pre-2016 or the post-2016 updated version of the Nutrition Facts Label by clicking on the image of each food product.

2025-07-21

2025-11-30

We examine the effect of nutrition information and the updated NFL on three outcomes:
1) Nutrition literacy, measured by the accuracy of ranking products by healthfulness based on objective nutrition scores;
2) Healthfulness of preferences, assessed by alignment between participants’ preference rankings and rankings derived from objective nutrition scores; and
3) Literacy–preference gap, defined as the discrepancy between nutrition literacy and the healthfulness of preferences.

We assess participants’ perceived healthfulness and preferences for food products by asking them to rank five items within each of the three categories: savory, sweet, and beverage. Separate ranking tasks are conducted for perceived healthfulness and for preference, and these are repeated across all three categories. Then, We construct measures of nutrition literacy, healthfulness of preferences, and literacy–preference gap by using root mean squared errors.

General nutrition knowledge

Participants answer twenty true-or-false questions based on Dickson-Spillmann et al. (2011), with each correct response earning one point. The total score reflects their baseline nutrition knowledge. Additionally, we evaluate participants’ understanding of Nutrition Facts Labels using seven questions adapted from Kim et al. (2021, Table 1).

This study is a one-time online survey conducted via the YouGov platform. We recruit 2,000 adults (ages 18 and older) to investigate how nutrition information and changes to the Nutrition Facts Label affect participants’ perceived healthfulness and preference rankings of food products. Participants complete two separate ranking tasks: a healthfulness ranking and a preference ranking. In both tasks, they evaluate five products in each of three categories: savory, sweet, and beverage.
For the healthfulness ranking, participants are asked to rank the products in each category from the healthiest to the least healthy. To encourage careful decision-making, participants receive monetary incentives based on the accuracy of their healthfulness rankings. They can earn up to $7.50 in additional compensation, on top of the base participation reward, with higher earnings awarded for more accurate rankings.
For the preference ranking task, participants rank the products in each category from the one they most prefer to the one they least prefer. To incentivize truthful reporting of preferences, 100 participants are randomly selected to receive a home-delivered food and beverage basket. The number of items received per category depends on a randomly drawn number from 1 to 5: selected participants receive the products they ranked equal to or higher than the drawn number in each category.
The sequence of healthfulness and preference ranking tasks, as well as the presentation order of categories and food products within each category, are fully randomized across participants.
Participants are randomly assigned to one of the four experimental groups following a 2 × 2 design: 1) Nutrition Information (provided vs. not provided), and 2) Nutrition Facts Label (pre-2016 vs. post-2016). In addition, we examine whether treatment effects differ by participants’ characteristics, especially socioeconomic status, which is not experimentally assigned but measured using background demographic data (education and household income) collected by YouGov.

The study proceeds as follows:
1. Baseline Assessment: All participants first answer a series of general nutrition knowledge questions.
2. Information Treatment (if assigned): Participants in the information treatment group receive general nutrition guidance. To ensure engagement, 1) a minimum viewing time is enforced before participants can proceed, and 2) simple comprehension check questions follow.
3. Product Ranking Tasks: Participants evaluate three categories of food and beverage products: savory, sweet, and beverage. Each category contains five products that vary in healthfulness, as defined by the Guiding Stars nutrition rating system. Participants complete two ranking tasks, 1) perceived healthfulness ranking, and 2) personal preference ranking. The order of the categories, the order of products within each category, and the sequence of the two ranking tasks is randomized across participants. During these tasks, participants are able to view either the pre-2016 or post-2016 updated Nutrition Facts Label by clicking on the image of the product, depending on their assigned condition.
4. Exclusion Preferences: After completing the preference rankings, participants are asked to indicate whether there are any products they would prefer not to receive. This ensures that, if selected to receive a food and beverage basket, they are only sent preferred items.
5. Survey Questionnaire: At the end of the study, participants complete a questionnaire to collect data on household characteristics, grocery shopping habits, health- and diet-related issues, and NFL reading habits.

Participants are randomly assigned to experimental conditions using YouGov’s proprietary randomization platform.

The unit of randomization is the individual participant.

No

Not applicable – Treatments are individually randomized; no clustering was used.

2,000 participants

Participants are randomly assigned to one of four experimental conditions based on a 2 × 2 factorial design:
- General Nutrition Information: Information vs. No Information
- Nutrition Facts Label (NFL): Old NFL vs. New NFL
This results in four treatment arms, with approximately 500 participants per arm, for a total sample size of 2,000 participants.
We also examine heterogeneous treatment effects by participant characteristics including socioeconomic status, which is not experimentally assigned but measured based on participants’ demographic information (household income and education) provided by YouGov.