Designing and implementing an experimental survey on knowledge and perceptions about alcohol warning labels

Objectives: This paper describes the design and implementation of an online survey experiment to investigate the effects of alcohol warning labels on alcohol-related knowledge, risk perceptions and intentions. Method: The survey collected self-reported data from 14 European countries through two waves of data collection with different recruitment strategies: dissemination via social media and public health agencies was followed by paid-for Facebook ads. The latter strategy was adopted to achieve broader population representation. Post-stratification weighting was used to match the sample to population demographics. Results: The survey received over 34,000 visits and resulted in a sample size of 19,601 participants with complete data on key sociodemographic characteristics. The responses in the first wave were over-representing females and higher educated people, thus the dissemination was complemented by the paid-for Facebook ads targeting more diverse populations but had higher attrition rate. Conclusion: Experiments can be integrated into general population surveys. Pan-European results can be achieved with limited resources and a combination of sampling methods to compensate for different biases, and statistical adjustments.