The COVID-19 pandemic had a huge impact on people’s use of public transport during 2020. While this was largely due to changes in working and social patterns that were imposed on the UK, many people still needed to travel. It was particularly vital that key workers were not deterred by fears about catching the virus, but the need to support and maintain sustainable travel choices meant that wider confidence was also important.
Wearing face coverings in public settings had become established as a vital factor in minimising the risk of spreading the virus, and the transport network, with its enclosed spaces and potential for bringing very different types of people together, was a perfect example of a risky environment. It was critical that people complied with the rules around wearing face coverings whilst travelling both to reduce the risk of transmission and to bolster confidence in the safety of the network.
The Department for Transport (DfT) in the UK wanted to be able to provide consistent advice to Network Rail and train operating companies on communications that would increase confidence in the network (and therefore greater usage of it), and greater correct usage of face coverings, without marginalising those who were exempt. The Department needed evidence of what works quickly to inform ongoing decisions and dialogue with these delivery partners.
We conducted a 6-arm online randomised controlled trial (RCT) with 3,000 participants, testing the effectiveness of five posters at improving compliance with the wearing of face coverings, compared to a control with no intervention.
We simulated a multi-modal journey using photographs to depict key stages, including bus, train and walking. In some photographs, we inserted the material being tested in each trial arm in a natural location (e.g. inside a train carriage, see right). The simulation immersed participants in the journey and permitted natural exposure to the materials. It also allowed them to take real decisions throughout their journey: they chose whether or not to wear a face covering at each stage, using a series of icons to depict their ‘mask status’.
The primary outcome was the proportion of participants who choose to wear a face covering correctly at appropriate journey points.
There was lower compliance for a poster that focused on enforcement and a fine (7ppt compared to the control, p<0.05). We were therefore able to recommend against using this approach.
There were no significant positive results on the primary outcome, but a poster using social proof and a ‘thank you’ performed best on almost every measure of compliance, as well as secondary outcome measures around confidence in using public transport. This poster was subsequently used across the network, on DfT’s recommendation.
Compliance in our simulated environment was similar to that observed in the field (~90%), and it fluctuated plausibly across different stages of the journey (e.g. it was lower when outside a railway station), which lent confidence to our methodology and results. The study also demonstrated the value of including secondary measures to provide evidence that can inform recommendations in the case of a null result on the primary measure.