Developing and Evaluating Product Messaging

Purpose: To assess the impact of taxes, warnings, and a combination of taxes and warnings on US adults' decisions to purchase products that contain red meat in an online grocery store. Procedures (methods): Participants will be recruited from Prime Panels (an online panel research company). Following online consent, participants will be assigned to one of four trial arms: 1) Control (no warning and no tax), 2) Warnings (all products that contain red meat have a health warning and environmental warning), 3) Tax (30% tax on products that contain red meat), and 4) Combined warning and tax (all products that contain red meat will have the two warnings and a 30% tax). Then, participant will enter an online grocery store reflecting their assigned arm. The participant will be instructed to complete a shopping task in the online grocery store. After completing the shopping task, participants will be redirected to an online survey and answer a series of questions about the shopping task, labels (excluding tax and control groups), and taxes (excluding warning and control groups). Questions will also include standard demographic and health related variables.

Participants will be recruited from Prime Panels (an online panel research company). Following online consent, participants will be randomly assigned to one of four trial arms affecting the appearance of an online grocery store. Once in the online grocery store, participants will complete a shopping task. The participants will have 9 specified items to add to their shopping cart (1 pizza, 1 burrito, Burger patties [meat or vegetarian], Breakfast sausages [meat or vegetarian], 1 frozen individual meal, 1 loaf of bread, 1 sandwich filling [for example, ham, turkey, or peanut butter], 1 pack of tortillas, 1 taco filling [for example, steak, chicken, or beans]). Participants will be informed that they will not be spending any of their own money to complete the shopping task. After completing the shopping task, participants will be redirected to a survey. They will provide self-reported responses to survey questions (e.g. perceived healthfulness from eating red meat, perceived risk of cancer from eating red meat, thinking about the health/environmental harms of food products). Participants will also answer demographic and health related questions.

The warning labels are black octagons with white text that appear next to images of products that contain red meat. One label says "WARNING: Eating red meat contributes to colon and rectal cancer" and the other label says "WARNING: Eating red meat harms the environment."

The Combined Warning Labels and Tax arm features both the warning labels and tax. The warning labels are black octagons with white text that appear next to images of products that contain red meat. One label says "WARNING: Eating red meat contributes to colon and rectal cancer" and the other label says "WARNING: Eating red meat harms the environment." The tax is a 30% increase in the price of red meat products.

Participants will complete a shopping task in the online grocery store, and the products that contain red meat will not have warning labels or a 30% increase in price.

While completing the shopping task in the online grocery store, the health and environmental warning labels will appear next to the image of every product that contains red meat.

While completing the shopping task in the online grocery store, all products containing red meat will have a 30% increase in price.

While completing the shopping task in the online grocery store, all products containing red meat will have the health and environmental warning labels appear next to the product image and a 30% increase in price.

Total grams of saturated fat of products in a participant's shopping basket. This will be calculated by adding together the total saturated fat from each product in the participant's shopping basket. Saturated fat for each product will come from the product's nutrition facts panel. All products in the store are real life products that have designated nutrition facts panels.

Total grams of sodium of products in a participant's shopping basket. This will be calculated by adding together the total sodium from each product in the participant's shopping basket. Sodium for each product will come from the product's nutrition facts panel. All products in the store are real life products that have designated nutrition facts panels.

Total calories of products in a participant's shopping basket. This will be calculated by adding together the total calories from each product in the participant's shopping basket. Calories for each product will come from the product's nutrition facts panel. All products in the store are real life products that have designated nutrition facts panels.

Perceived healthfulness of eating red meat. Likert response options are on a 1 to 5 scale, with higher scores representing a higher perceived healthfulness of eating red meat.

Perceived risk of colon and rectal cancer from eating red meat. Likert response options are on a 1 to 5 scale, with higher scores representing a higher perceived risk of colon and rectal cancer from eating red meat.

Perceived risk of environmental harms from eating red meat. Likert response options are on a 1 to 5 scale, with higher scores representing a higher perceived risk of environmental harms from eating red meat.

Thinking about the health harms of food products while completing the shopping task. Likert response options are on a 1 to 5 scale, with higher scores representing a higher amount of thinking about health harms of food products.

Thinking about the environmental harms of food products while completing the shopping task. Likert response options are on a 1 to 5 scale, with higher scores representing a higher amount of thinking about the environmental harms of food products.

Thinking about the price of food products while completing the shopping task. Likert response options are on a 1 to 5 scale, with higher scores representing a higher amount of thinking about the price of food products.

Perceived healthfulness of specific red meat products (burgers, pepperoni pizza, and deli ham). Likert response options are on a 1 to 5 scale, with lower scores representing worse for health and higher scores representing better for health.

Perceived environmental harm of specific red meat products (burgers, pepperoni pizza, and deli ham). Likert response options are on a 1 to 5 scale, with lower scores representing worse for the environment and higher scores representing better for the environment.

Perceived cost of the burger product. Likert response options are on a 1 to 5 scale, with lower scores representing less expensive and higher scores representing more expensive.

Perceived cost of the pizza product. Likert response options are on a 1 to 5 scale, with lower scores representing less expensive and higher scores representing more expensive.

Perceived cost of the ham product. Likert response options are on a 1 to 5 scale, with lower scores representing less expensive and higher scores representing more expensive.

Intention to reduce red meat consumption in the next 30 days. Likert response options are on a 1 to 5 scale, with higher scores representing a higher intention to reduce red meat consumption in the next 30 days.

Support of policy for taxes on red meat products in the U.S. Likert response options are on a 1 to 5 scale, with higher scores representing a greater deal of support. Those who indicated '4' or '5' were assumed to more strongly agree to support a red meat product tax. This measure reports the percentage of those with a response of '4' or '5'.

Support of policy for health warning labels on red meat products in the U.S. Likert response options are on a 1 to 5 scale, with higher scores representing a greater deal of support. Those who indicated '4' or '5' were assumed to more strongly agree to support a health warning label. This measure reports the percentage of those with a response of '4' or '5'.

Support of policy for environmental warning labels on red meat products in the U.S. Likert response options are on a 1 to 5 scale, with higher scores representing a greater deal of support. Support of policy for taxes on environmental warning labels in the U.S. Likert response options are on a 1 to 5 scale, with higher scores representing a greater deal of support. Those who indicated '4' or '5' were assumed to more strongly agree to support an environmental warning label. This measure reports the percentage of participants with a response of '4' or '5'.

Acceptability of the online grocery store. Likert response options are on a 1 to 5 scale, with higher scores representing a higher amount of agreement. Acceptability of the online grocery store is measured using 3 items adapted from Waterlander et al., 2011. Responses to the 3 items will be averaged to create an acceptability score sum ranging from 3-15 then divided by the total number of questions asked (n = 3). Prompt: Say whether you agree or disagree with the following statements. I could easily find all of the food and beverages I was looking for in the online grocery store. There were enough food and beverage options in the online grocery store. This online grocery store felt like a real online grocery store. Waterlander WE, Scarpa M, Lentz D, Steenhuis IHM. The virtual supermarket: an innovative research tool to study consumer food purchasing behavior. 2011;11:589. DOI: 10.1186/1471-2458-11-589.

Ease of use of the online grocery store. Likert response options are on a 1 to 5 scale, with higher scores representing more agreement that the online grocery store is easy to use.

Inclusion Criteria: 18 years or older Consumes red meat at least once per week. Responsible for at least 50% of grocery shopping for the household. Lives in the United States Exclusion Criteria: Individuals who have participated in research studies from previous aims of this research.

A deidentified dataset and the survey codebook will be published to the Harvard Dataverse. The study protocol and statistical analysis plan will be published as supporting information with the study results in a peer-reviewed journal.

The deidentified dataset and corresponding survey codebook will be available upon publication of study results and will remain on the Harvard Dataverse indefinitely.