Alcohol Health Warning Study
Principal Investigator(s): View help for Principal Investigator(s) Emily Brennan, Centre for Behavioural Research in Cancer, Cancer Council Victoria
Version: View help for Version V1
| Name | File Type | Size | Last Modified |
|---|---|---|---|
|
|
application/x-stata-dta | 314.6 KB | 10/05/2022 10:18:PM |
Project Citation:
Brennan, Emily. Alcohol Health Warning Study. Ann Arbor, MI: Inter-university Consortium for Political and Social Research [distributor], 2022-10-12. https://doi.org/10.3886/E175721V1
Project Description
Summary:
View help for Summary
Health warning labels (HWLs) on alcohol containers may help reduce population-level alcohol consumption. However, few studies have examined the most effective formats for alcohol HWLs. This study tested the effects of three different types of alcohol HWLs. We conducted a between-subjects online experiment among N= 1,755 Australian adult drinkers with an initial exposure session, measurement of immediate post-exposure outcomes, repeated exposure over the subsequent eight days and measurement of additional outcomes at follow-up. Australian adults were randomised to one of five conditions: (a) No HWL control; (b) DrinkWise control; (c) Text-Only intervention (Text-Only); (d) Text-and-Pictogram intervention (Text+Pictogram); or (e) Text-and-Photograph intervention (Text+Photograph). In each intervention condition, participants were exposed to eight HWLs, each depicting a different alcohol-related harm.
Funding Sources:
View help for Funding Sources
Australian National Health and Medical Research Council (1129002)
Scope of Project
Subject Terms:
View help for Subject Terms
alcohol;
alcohol-related harm;
cancer;
knowledge;
intentions;
public health;
warning labels
Geographic Coverage:
View help for Geographic Coverage
Australia
Time Period(s):
View help for Time Period(s)
3/12/2020 – 4/8/2020
Collection Date(s):
View help for Collection Date(s)
3/12/2020 – 4/8/2020
Universe:
View help for Universe
Australian adults aged 18-69 years who consumed alcohol on average at least weekly during the past year.
Data Type(s):
View help for Data Type(s)
experimental data
Collection Notes:
View help for Collection Notes
Recruitment and data collection commenced on 12 March 2020. Due to the rapidly changing environment and widespread uncertainty caused by COVID-19, recruitment ceased prematurely on 25 March 2020 and all follow-up data collection was completed by 8 April 2020.
Methodology
Response Rate:
View help for Response Rate
In total, N=1,829 participants were recruited, but dropped out after randomisation, leaving N=1,755 (67.5% of the planned sample size) for analyses of outcomes measured immediately post-exposure. Of the N=1,755 participants who completed the baseline survey and initial exposure session, 668 (38.1%) did not complete follow-up, leaving N=1,087 participants for analyses using follow-up outcomes (59.7% of the planned sample size at follow-up). The attrition rate was similar across conditions (χ2(4)=1.65; p=0.799). There were no imbalances across conditions in the characteristics of participants retained at follow-up.
Sampling:
View help for Sampling
Participants were recruited through an online non-probability panel accredited under the International Organization for Standardization’s standards for Market, Opinion and Social Research (AS ISO 20252), where participants opt-in to receive email invitations to participate in market research. Survey participants receive points they may accrue and redeem for gift vouchers. Panel members received an email invitation, clicked through to the study, completed screening questions and then implied consent by clicking through to further questions assessing alcohol consumption, after which they were randomised to experimental conditions. Quotas were applied to achieve approximately even numbers by gender and proportional quotas for age (18-29, 30-49 and 50-69 years) based on the distribution of weekly drinkers aged 18-69 years in 2019.
Data Source:
View help for Data Source
Experimental data collected through a web-based survey.
Collection Mode(s):
View help for Collection Mode(s)
web-based survey
Scales:
View help for Scales
Demographic characteristics
Participants reported age, gender, highest level of educational attainment and if they were a parent or guardian. Socioeconomic status was determined using participants’ postcode and an Index of Relative Socio-Economic Disadvantage.
Drinking characteristics
Participants’ past 12 month usual pattern of alcohol consumption was measured using the graduated quantity–frequency measure. Participants were provided with a visual guide of the number of standard drinks (one standard drink in Australia = 10g alcohol) in common servings of different alcohol types. Applying the Australian National Health and Medical Research Council’s 2009 Low-Risk Drinking Guidelines, current at the time of the study, participants were classified as at high risk of long-term harm (LTH) if they consumed > two drinks per day on average and/or at high risk of short-term harm (STH) if they had > four drinks on any occasion at least once a month. Participants also indicated the perceived amount of alcohol they currently drink, with ‘self-perceived high-risk drinkers’ classified as those who responded that “I definitely drink more than I should” or “I probably drink more than I should”, and ‘self-perceived low-risk drinkers’ as those who responded with either “the amount I drink is ok” or “I could drink more than I do”.
Intentions to reduce drinking
Immediately post exposure, participants indicated on an ordinal scale (1 = “definitely will not”, 2 = “probably will not”, 3 = “probably will”, and 4 = “definitely will”) the extent to which “In the next week, will you try to drink less alcohol”, and the extent to which they will in the next month, “reduce how often you drink alcohol”, “reduce the amount of alcohol you have on each drinking occasion,” and “avoid drinking alcohol completely”. Responses were dichotomised and classified as ‘definitely/probably will not’ or ‘definitely/probably will’. Since the items “reduce how often you drink alcohol” and “reduce the amount of alcohol you have on each drinking occasion’’ were highly correlated (polychoric correlation ρ = .72), they were combined, with those who answered ‘definitely/probably will’ for either or both statements being classified as ‘definitely/probably will’ in the combined variable.
Past week alcohol consumption was assessed at follow-up using the seven-day follow-back measure.
Intentions to reduce drinking in the next week and month at follow-up were measured using the same questions as those completed immediately post-exposure.
Readiness to change
We used the readiness to change ruler for decreased drinking at follow-up. Participants indicated their readiness to reduce their drinking using an 11-point scale (with the anchor points of 0 = “I never think about drinking less”, 3 = “sometimes I think about drinking less”, 5 = “I have decided to drink less”, 7 = “I am already trying to cut back on my drinking”, and 10 = “my drinking has changed, I now drink less than before”).
Frequency of thinking about alcohol-related health risks in the past week was adapted from tobacco HWL studies and assessed at follow-up. Participants were asked “In the past week, to what extent, if at all, did you think about the health risks associated with drinking alcohol?”. Responses were assessed using a four-point scale (with the option of selecting “don’t know”) and dichotomised for analysis (‘somewhat/a lot’ or ‘not at all/a little/don’t know’).
Awareness of alcohol-related health harms
Participants were asked if “Consuming alcohol would”: (i) “increase your risk of cancer”; (ii) “increase your risk of liver damage”; and (iii) “increase your risk of heart disease” at follow-up. Given that the DrinkWise labels provided the message that ‘It is safest not to drink during pregnancy’, participants were also asked if “Consuming alcohol would”: (iv) “increase the risk of a woman experiencing pregnancy complications”. Responses were assessed on a seven-point scale and dichotomised for analysis (‘slightly agree/agree/strongly agree’ or ‘strongly disagree/disagree/slightly disagree/neither agree nor disagree’).
Negative emotional arousal
At follow-up, participants were asked: “Thinking about the images of alcohol containers I’ve seen as part of this study, I felt: disgusted / afraid / uncomfortable / worried”. Responses were rated on seven-point scales with three anchor points (1 = “not at all”, 4 = “moderately”, and 7 = “very”). A scale indicating participants’ negative emotional arousal was created by taking the mean of the four items (Cronbach’s α = .92).
Positive emotional arousal
At follow-up, participants were asked: “Thinking about the images of alcohol containers I’ve seen as part of this study, I felt: excited / pleased”. Responses were rated on seven-point scales with three anchor points (1 = “not at all”, 4 = “moderately”, and 7 = “very”). A scale indicating participants’ positive emotional arousal was created by taking the mean of the two items (Cronbach’s α = .85).
Show/mention images to others
At follow-up, participants were asked: (i) “As part of this project, you received emails over the past eight days to view images of different alcoholic drinks. Did you show any of the images to other people?” and (ii) “Did you talk about the images with anyone, even if you didn’t show them an image?”. Response options were ‘yes’ and ‘no’. Responses were combined so that if the participant answered yes to one or both items, they were coded as ‘Yes, shared/talked’ versus ‘No, did not share/talk’.
Participants reported age, gender, highest level of educational attainment and if they were a parent or guardian. Socioeconomic status was determined using participants’ postcode and an Index of Relative Socio-Economic Disadvantage.
Drinking characteristics
Participants’ past 12 month usual pattern of alcohol consumption was measured using the graduated quantity–frequency measure. Participants were provided with a visual guide of the number of standard drinks (one standard drink in Australia = 10g alcohol) in common servings of different alcohol types. Applying the Australian National Health and Medical Research Council’s 2009 Low-Risk Drinking Guidelines, current at the time of the study, participants were classified as at high risk of long-term harm (LTH) if they consumed > two drinks per day on average and/or at high risk of short-term harm (STH) if they had > four drinks on any occasion at least once a month. Participants also indicated the perceived amount of alcohol they currently drink, with ‘self-perceived high-risk drinkers’ classified as those who responded that “I definitely drink more than I should” or “I probably drink more than I should”, and ‘self-perceived low-risk drinkers’ as those who responded with either “the amount I drink is ok” or “I could drink more than I do”.
Intentions to reduce drinking
Immediately post exposure, participants indicated on an ordinal scale (1 = “definitely will not”, 2 = “probably will not”, 3 = “probably will”, and 4 = “definitely will”) the extent to which “In the next week, will you try to drink less alcohol”, and the extent to which they will in the next month, “reduce how often you drink alcohol”, “reduce the amount of alcohol you have on each drinking occasion,” and “avoid drinking alcohol completely”. Responses were dichotomised and classified as ‘definitely/probably will not’ or ‘definitely/probably will’. Since the items “reduce how often you drink alcohol” and “reduce the amount of alcohol you have on each drinking occasion’’ were highly correlated (polychoric correlation ρ = .72), they were combined, with those who answered ‘definitely/probably will’ for either or both statements being classified as ‘definitely/probably will’ in the combined variable.
Past week alcohol consumption was assessed at follow-up using the seven-day follow-back measure.
Intentions to reduce drinking in the next week and month at follow-up were measured using the same questions as those completed immediately post-exposure.
Readiness to change
We used the readiness to change ruler for decreased drinking at follow-up. Participants indicated their readiness to reduce their drinking using an 11-point scale (with the anchor points of 0 = “I never think about drinking less”, 3 = “sometimes I think about drinking less”, 5 = “I have decided to drink less”, 7 = “I am already trying to cut back on my drinking”, and 10 = “my drinking has changed, I now drink less than before”).
Frequency of thinking about alcohol-related health risks in the past week was adapted from tobacco HWL studies and assessed at follow-up. Participants were asked “In the past week, to what extent, if at all, did you think about the health risks associated with drinking alcohol?”. Responses were assessed using a four-point scale (with the option of selecting “don’t know”) and dichotomised for analysis (‘somewhat/a lot’ or ‘not at all/a little/don’t know’).
Awareness of alcohol-related health harms
Participants were asked if “Consuming alcohol would”: (i) “increase your risk of cancer”; (ii) “increase your risk of liver damage”; and (iii) “increase your risk of heart disease” at follow-up. Given that the DrinkWise labels provided the message that ‘It is safest not to drink during pregnancy’, participants were also asked if “Consuming alcohol would”: (iv) “increase the risk of a woman experiencing pregnancy complications”. Responses were assessed on a seven-point scale and dichotomised for analysis (‘slightly agree/agree/strongly agree’ or ‘strongly disagree/disagree/slightly disagree/neither agree nor disagree’).
Negative emotional arousal
At follow-up, participants were asked: “Thinking about the images of alcohol containers I’ve seen as part of this study, I felt: disgusted / afraid / uncomfortable / worried”. Responses were rated on seven-point scales with three anchor points (1 = “not at all”, 4 = “moderately”, and 7 = “very”). A scale indicating participants’ negative emotional arousal was created by taking the mean of the four items (Cronbach’s α = .92).
Positive emotional arousal
At follow-up, participants were asked: “Thinking about the images of alcohol containers I’ve seen as part of this study, I felt: excited / pleased”. Responses were rated on seven-point scales with three anchor points (1 = “not at all”, 4 = “moderately”, and 7 = “very”). A scale indicating participants’ positive emotional arousal was created by taking the mean of the two items (Cronbach’s α = .85).
Show/mention images to others
At follow-up, participants were asked: (i) “As part of this project, you received emails over the past eight days to view images of different alcoholic drinks. Did you show any of the images to other people?” and (ii) “Did you talk about the images with anyone, even if you didn’t show them an image?”. Response options were ‘yes’ and ‘no’. Responses were combined so that if the participant answered yes to one or both items, they were coded as ‘Yes, shared/talked’ versus ‘No, did not share/talk’.
Weights:
View help for Weights
No weighting variables are included in the data file and weights were not used during data analysis.
Unit(s) of Observation:
View help for Unit(s) of Observation
Rating scores
Geographic Unit:
View help for Geographic Unit
Postcode
Related Publications
Published Versions
Report a Problem
Found a serious problem with the data, such as disclosure risk or copyrighted content? Let us know.
This material is distributed exactly as received from the data depositor. As of April 2026, depositors are required to submit study materials in accessible formats. ICPSR has not reviewed, checked, or processed this material. For additional information about the study, please contact the investigator(s) directly. If you have questions about the accessibility of materials distributed by ICPSR or require further assistance, please visit ICPSR's Accessibility Center.