Alcohol and Cancer Message Testing Experiment

Exposure to two alcohol-cancer messages that convey the probability of experiencing a health harm (i.e., cancer) from a particular hazard (i.e., drinking alcohol).

The intervention will be administered in an online survey experiment as static images displayed to participants.

Randomized exposure to one of five alcohol-cancer health campaign message frames, or a non-cancer alcohol harms frame.

Exposure to two alcohol-cancer messages that encourage reducing how much alcohol people drink. The intervention will be administered in an online survey experiment as static images displayed to participants.

Randomized exposure to one of five alcohol-cancer health campaign message frames, or a non-cancer alcohol harms frame.

Exposure to two alcohol-cancer messages that convey how serious getting alcohol-associated cancer will be. The intervention will be administered in an online survey experiment as static images displayed to participants.

Randomized exposure to one of five alcohol-cancer health campaign message frames, or a non-cancer alcohol harms frame.

Exposure to two alcohol-cancer messages that emphasize the effectiveness of protective behaviors for reducing alcohol-associated cancer risk. The intervention will be administered in an online survey experiment as static images displayed to participants.

Randomized exposure to one of five alcohol-cancer health campaign message frames, or a non-cancer alcohol harms frame.

Exposure to two alcohol-cancer messages that emphasize completely abstaining from drinking alcohol. The intervention will be administered in an online survey experiment as static images displayed to participants.

Randomized exposure to one of five alcohol-cancer health campaign message frames, or a non-cancer alcohol harms frame.

Exposure to two alcohol-cancer messages that do not discuss alcohol-associated cancer risk. Instead, the messages use language from the existing U.S. Surgeon General's warning on alcoholic beverages. The intervention will be administered in an online survey experiment as static images displayed to participants.

Randomized exposure to one of five alcohol-cancer health campaign message frames, or a non-cancer alcohol harms frame.

Participants randomized to control will not be exposed to any message stimuli.

A 3-item scale adapted from prior effects perception PME and substance use literature. After each message exposure, participants will be asked to respond to the following items: 1) This message discourages me from wanting to drink alcohol; 2) This message makes me concerned about the health effects of drinking alcohol; and 3) This message makes drinking alcohol seem unpleasant to me. Items will be assessed on a 5-point scale ranging from strongly disagree (1) to strongly agree (5). We will assess internal consistency using Cronbach's alpha. Assuming adequate measure reliability, we will average the items into a single composite PME measure.

Intentions will be assessed with a 3-item scale adapted from prior substance use and health communication research. Participants will answer each of the following questions: How likely is it that you will drink alcohol in the... 1) next week, 2) next month, and 3) next year. Responses will be on 5-point scales ranging from "not at all" (1) to "extremely" (5). We will assess internal consistency using Cronbach's alpha. Assuming adequate measure reliability, we will average the items into a single composite measure.

Message reactance will be assessed with a brief 3-item scale used in prior research. Participants will respond to the following statements: 1) This message is trying to manipulate me, 2) The health effect in this message is overblown, and 3) This message annoys me. Responses will be assessed on 5-point scales ranging from "strongly disagree" (1) to "strongly agree" (5). We will assess internal consistency using Cronbach's alpha. Assuming adequate measure reliability, we will average the items into a single composite measure.

A single item that has participants respond to the following statement: "This message makes me think drinking non-alcoholic beer is healthier than alcoholic beer." This item will be assessed on a 5-point scale ranging from strongly disagree (1) to strongly agree (5).

A 5-item scale assessing perceptions of each message, adapted from prior work. After each message exposure, participants will respond to the following items: 1) "This message grabs my attention"; 2) "This message is believable"; 3) "This message is credible"; 4) "This message is trustworthy"; and 5) "I learned something new from this message." Items will be assessed on a 5-point scale ranging from strongly disagree (1) to strongly agree (5). We will assess internal consistency using Cronbach's alpha. Assuming adequate measure reliability, we will average the items into a single composite measure.

A 3-item measure assessing attitudes toward drinking alcohol. Participants will rate drinking alcohol on three 5-point bipolar scales: 1) harmful/safe; 2) bad/good; and 3) unhealthy/healthy. We will assess internal consistency using Cronbach's alpha. Assuming adequate reliability, we will average the items into a single composite attitude measure.

Assessed only among participants who currently drink alcohol. Participants will be asked, "How likely are you to reduce how much alcohol you drink?" Responses will be assessed on a 5-point scale ranging from not at all (1) to very much (5).

Causal beliefs will be assessed with a single item: "How likely do you think it is that alcohol causes cancer?" Responses will be assessed on a 5-point scale ranging from not at all (1) to very much (5).

Participants who currently drink alcohol will be asked, "How likely are you to quit drinking alcohol?" Responses will be assessed on a 5-point scale ranging from not at all (1) to very much (5).

We will assess risk beliefs using a 3-item scale assessing affective, experiential, and deliberative perceptions of cancer risk among participants who currently drink alcohol. The measure stem will state, "If you continue to drink alcohol at your current level…" Participants will then answer the following items: 1) "How worried are you about eventually getting cancer?"; 2) "To what extent could you see yourself eventually getting cancer?"; and 3) "How likely do you think it is that you will eventually get cancer?" Items will be assessed on a 5-point scale ranging from not at all (1) to very much (5). We will assess these items separately, as they tap into different tri-risk constructs; however, for exploratory purposes we will also assess internal consistency of the three items using Cronbach's alpha. Assuming adequate reliability, we will average the items into a single composite measure.

Assessed with a single item: "I believe a message like this should be added to alcohol beverage containers." Responses will be assessed on a 5-point scale ranging from strongly disagree (1) to strongly agree (5).

Assessed with a single item: "Seeing a message like this on an alcoholic beverage would make me think twice about drinking alcohol." Responses will be assessed on a 5-point scale ranging from strongly disagree (1) to strongly agree (5).