EMA for the Exposure to Alcohol Advertisements

• Study design The proposed project is a combination of (1) a time-based system-triggered EMA, which will collect real-time exposure to alcohol advertisements and PADC, and (2) longitudinal surveys involving university students who are current alcohol users. Consented participants will be individually randomized to 2 groups - the EMA and non-EMA group. The EMA group will first complete a baseline questionnaire, followed by 5 regularly prompted EMA surveys via a smartphone application (App) every day for 14 consecutive days. The EMA includes questions of the self-reported exposure to and /or responses to alcohol advertising PADC. 2 weeks after the completion of the EMA, they will be contacted to complete a telephone follow-up survey. The non-EMA group will only complete the baseline and the follow-up surveys.
• Subjects The study targets to recruit 100 students (EMA: 50; non-EMA: 50) who are Hong Kong residents and are studying in undergraduate or postgraduate programmes in any Hong Kong tertiary institutions. Other inclusion criteria will include (1) aged between 18 and 35 years, (2) using a mobile smartphone with internet access, (3) drank any alcohol in past 30 days (current drinkers), (4) he/she will stay in Hong Kong during the 2-week EMA study period, and (5) able to read and write in Chinese.
• Procedures A mass email containing the study information will be sent to all undergraduate and postgraduate HKU students for recruitment. Also, subject recruitment ad will be posted in different online discussion boards. Voluntary participants can directly contact our recruitment staff via telephone, email or submission of online application form. The staff will assess the participants' eligibility by telephone. The staff will then randomise the eligible participants into either the EMA or non-EMA group and deliver consent forms to the participants face-to-face before conducting the baseline survey. Sequentially numbered, opaque sealed envelope (SNOSE) will be used for the random allocation. All participants will be invited to join a briefing session where the staff will administer the baseline survey. The staff will help the participants in the EMA group to install and set up the EMA App. The briefing will also emphasise the role of EMA participants that they should perform their normal activities as usual instead of finding out the alcohol advertisements and PADC intentionally. Within 2 weeks after completing the baseline survey, participants in EMA group will complete 5 EMAs prompted regularly every day for 14 consecutive days. The participants will then be contacted for a follow-up survey 2 weeks after the completion of EMA. Participants in non-EMA group will only complete a baseline questionnaire and a follow-up survey.
• Incentives for participants The EMA validity largely relies on the participants' compliance, therefore an attractive incentive is needed. All participants will be given a HK$100 gift voucher upon the completion of the baseline survey and 2-week follow-up survey. In addition, participants in EMA group will accumulate HK$15 gift voucher for completing 3 EMA surveys within each day in their account. If they complete all EMA surveys within a day, they can further accumulate HK$10 gift voucher. EMA participants who have successfully completed all EMAs in the 14-day period will be given additional $50 gift voucher.
• EMA app An EMA App will be developed to administer the regular prompts automatically. To protect the privacy of the participants, no personal information (e.g. names and telephone numbers) will be collected by the App. Only a 5-digit case number will be used for identification. Unless otherwise specified, the case number will be the last 5 digits of subject's phone number. Once the App is successfully installed, the user will be asked to choose a start date for the 14 consecutive days study period within the coming 2 weeks and set up awake time from 9:00am to 12:00nn in these study days to undertake the regularly-prompting EMA.

On each study day, the App will regularly prompt the participants to complete a questionnaire about their recent exposure to alcohol advertisements and PADC every 3 hours for 5 times. For example, if a participant sets the awake time to 9:00am, he will be asked to complete an EMA survey at 9:00am, 12:00nn, 3:00pm, 6:00pm and 9:00pm for the study period.

Each EMA survey will take about 1 minute to complete. If the user cannot respond to the EMA survey prompt immediately, he/she could choose to postpone the prompt for 5 minutes / 30 minutes. The EMA prompt will be re-generated to remind the user again. Each prompts not responded for 3 mins, or postponed more than once will be regarded as non-response.

All the above data, including the date and time of completing the EMA survey and responses to the prompts will be sent to our server immediately. Participants can check their progress and record about their EMA surveys completion status and the amount of rewarded gift voucher in the EMA App.

• Questionnaire design

• Baseline questionnaire There are 4 parts in the baseline questionnaire to investigate the participants' drinking history, their awareness of alcohol advertising and PADC, and their attitude towards alcohol drinking culture, alcohol advertisements and alcohol advertising and marketing policy.

Part 1 - Participants' history of alcohol consumption during the past year The Alcohol Use Disorder Identification Test (AUDIT) published by the World Health Organisation (WHO) will be taken as a reference to understand participants' history of hazard and harmful alcohol consumption, especially in excessive alcohol intake during the past year. The test consists of 3 stages, namely "Hazardous Alcohol Use", "Dependence Symptoms" and "Harmful Alcohol Use".

In the 1st stage, questions will be involving the frequency of drinking, typical quantity of drinking, and frequency of heavy drinking (>60 grams of pure alcohol intake on one occasion) to identify any sign of hazardous alcohol consumption. During the 2nd stage, subject's experience in impaired control over drinking, increased salience of drinking and morning drinking will be asked to evaluate subject's alcohol dependence. At the 3rd stage, subject's history of harmful alcohol use, including guilt after drinking, blackouts, alcohol-related injuries and others concerned about drinking will be asked. A final score based on the answers can be computed and be used to classify the subject into "Low-risk", "Risky", "Harmful", and "Almost certainly dependent".

Part 2 - Participants' current drinking habit and involvement in PADC events Subject's current drinking habit in the past month will be asked in details, including types of alcohol beverage consumed, frequency of unplanned drinking and participations of events promoting drinking culture.

Furthermore, the locations and activities where the subject has drunk any alcohol will be asked. Options for location include: accommodation, restaurant, campus, bar, public area, transportation, and rural outdoor areas. Whereas options for activities include: family meal gathering, business, university societies' event, high-table dinner, spectating sports/race, celebrating festivals.

In addition, experiences associated with passive and forced drinking will be asked, including negative influence by surrounding drinkers and situation where the subject had experienced forced drinking.

Part 3 - Participants' awareness of alcohol advertising and PADC To evaluate subjects' exposure to alcohol advertisements and PADC, we first separate the channel to traditional media (TV/ radio/ newspaper/ magazine), internet (website/ social media/ video platforms/ mobile applications) and other physical locations (outdoor billboard/ restaurant/ supermarket/ shopping mall/ concert/ sports event). After that, subjects' average daily view counts via the corresponding channel during the past week will be asked and a total view count will be computed for further analysis.

Part 4 - Participants' attitude towards alcohol consumption, drinking culture, alcohol advertisements, and alcohol related policy Subjects will be asked to rate their extent of agreement (from 1: strongly disagree to 5: strongly agree) with various views on alcohol drinking, such as its impact on social relationships, relaxation, personal popularity, and potential health benefits and risks. Subsequent questions will cover subject's view on drinking-culture promoting events/ festivals, effect of ads on subjects' drinking behaviour, followed by subject's view on advertisement of alcoholic products featuring food/ sports/ music, and subject's view on banning alcohol advertisements and PADC.

• EMA Surveys

In each EMA prompt, participants will be asked if they have seen any alcohol advertisements and PADC in the past 3 hours (except for the first EMA survey of the day, which asks about the period starting from the last EMA on the previous day rather than the past 3 hours). Based on the reported number of advertisement(s) and PADC, they have to indicate where the advertisement and/or PADC has appeared, such as via television, prints, radio, social media, webpage, public area etc. In each reported exposure, participants will be asked what type of alcoholic drink has been advertised, with options "Beer", "Wine", "Liquor/ Spirit", "Chinese spirit", "Cocktail", "Japanese Sake", "Alcopop", and "Others". Then, they will be asked 7 follow-up questions to assess his/her appraisals on the advertisement(s):

• The main actor/actress in the alcohol ads are attractive,
• The ad is innovative,
• I love the ad
• The ads effectively appealed me to purchase the advertised product(s).
• The ads effectively appealed me to join the PADC event(s), and
• I want to drink alcohol after watching the ads.
• I want to purchase alcohol after watching the ads

In each statement, participants will be asked to rate his/her agreement from "1 (strongly disagree)" to "5 (strongly agree)". Questions (1) to (3) will be asked for each advertisement the subject has seen. Questions (4) to (10) will be asked only once on each EMA day as an overall assessment for the subjects' response to these exposure.

Participants will also be prompted to report the amount and type of alcohol consumed in the past 3 hours, with options "Beer", "Wine", "Liquor/ Spirit", "Chinese spirit", "Cocktail", "Japanese Sake", "Alcopop" and "Others", whether they have purchased any alcohol and/or participated in any activities about promoting alcohol drinking culture.

In addition to the system-triggered EMAs, participants can take photographs of the alcohol advertisements they have seen anytime and upload them via the EMA App. These photographs will be used for a content analysis.

• Follow-up questionnaire The questions in the follow-up survey will be very similar to the baseline questionnaire, but mainly focusing on exposure to alcohol ads and PADC for the sake of comparison between the EMA group and non-EMA group. In addition, a section is added for participants to express their satisfaction or dissatisfaction regarding different aspect of the study, including the baseline survey, mobile application, EMA surveys, follow-up survey and the overall rating of the study. The user feedback will be used to improve the user experiences and effectiveness of similar research projects in the future.
• Statistical analysis Descriptive statistics about the exposure to the alcohol advertisements/ PADC will be computed. Multiple comparison of Pearson and Spearman correlation coefficient with Bonferroni correction will serve as a preliminary measure of consistency of individual participant's responses on baseline, EMA (if applicable) and follow-up surveys.

All daily observations are nested within each participant. We will examine the relationship between number of daily exposure and the binary alcohol consumption/ binge drinking (any alcohol consumption/ binge drinking during that day) with logistic regression model and generalised estimation equation model.

In the participant-level analysis, the reported weekly exposure count will be regressed on age, gender, programme enrolled and AUDIT test score with zero-inflated negative binomial regression model. This regression model is preferred over the Poisson regression model because some participants will have zero exposure for a period of time and there may be over-dispersion of the count outcomes (i.e. when variance is greater than the mean). While the association between exposure and other attitude subscales in the baseline survey will be analysed with ordered logistic regression.

The retrospective recall of exposure to alcohol advertisements assessed at the follow-up survey between the EMA and non-EMA groups will be compared using t-test if the count data distribution does not significantly deviate from normal distribution, otherwise Mann-Whitney U test will be applied.

In order to test for effect of EMA on exposure count, within the EMA group, we will compare the reported weekly exposure in baseline, EMA and the follow-up survey, repeated measures ANOVA or Friedman test will be applied, depending on whether the normality assumption has been met. Post hoc test such as paired t-test or Wilcoxon signed-rank test will then be applied to discover whether the EMA effect on exposure is positive compared to baseline/ follow-up.

Inclusion criteria

(1) aged between 18 and 35 years, (2) using a mobile smartphone with internet access, (3) self-report drink any alcohol in past 30 days (current drinkers), (4) he/she will stay in Hong Kong during the 2-week EMA study period, and (5) able to read and write in Chinese