This article was written by Peter Slattery, Research Affiliate, MIT FutureTech, MIT.

  • The problem: There was a lack of clarity on who needed to change what behaviours to address the Covid-19 pandemic in 2020.
  • Why it matters: Understanding public behaviours and attitudes through "living surveys" allowed policymakers to make data-driven decisions during the pandemic and could benefit governments addressing other important social issues.
  • The solution: Governments should design, pilot, test and scale up "living surveys" to cost-effectively gain insights into public behaviours and attitudes on relevant issues.

In March 2020, my colleagues at Ready Research and I started to wonder how we could help with the encroaching pandemic.

At the time, there was very little clarity on the answer to a critical question: ‘Who needs to do what differently?’.

State and federal governments, friends and family and news media seemed to be urgently communicating information to try and change health behaviours such as hand-washing and keeping physical distance.

But _who _needed to change their behaviour? Who wasn’t doing key protective behaviours? Was it everyone, mostly men, or mostly people in certain suburbs? Maybe people were already doing what was recommended, and we were wasting energy by repeatedly asking for it?

We started to wonder: what could social science researchers do to help?

💬 Let the author know your thoughts on using living surveys to better understand public behaviour by leaving a comment below

Eventually, this exploration converged into the SCRUB project, which aimed to provide current and future policymakers with actionable insights into public attitudes and behaviours relating to the Covid-19 pandemic.

Our major contribution was to develop a “living survey” — with both repeated cross-sectional and longitudinal sampling — that we could use to track relevant behaviours (e.g., handwashing and physical distancing), their variations by demographic and location and their determinants.

Where possible, we tested batteries of messaging interventions during waves of data collection.

Although we collected international data and engaged with researchers and officials in a range of places, including Peru and Saudi Arabia, the majority of our work focused on understanding Australian behaviour.

This was done in partnership with BehaviourWorks Australia (at Monash Sustainable Development Institute in Monash University) and Australian Catholic University and was supported by more than 100 international researchers.

We collected information and tested interventions every 3–4 weeks from March 2020 until June 2021. For instance, we measured:

  • What people were doing to keep themselves safe (e.g., distancing, mask-wearing, willingness to vaccinate, compliance with rules) and why;
  • The impact of Covid-19 on lives and livelihoods (e.g., mental health and wellbeing, work arrangements, government payments and community services, worries);
  • People’s beliefs and expectations of a sustainable recovery (e.g., household finances, travel, Covid-19-safe workplaces);
  • Changes in related attitudes related issues such as meat consumption and pandemic preparedness.

After each wave of data collection, we generated and disseminated a report for relevant policymakers.

So how did it go?

By the conclusion of the project we had received more than 50,000 surveys from more than 40 countries (including around 3000 international responses).

Our data served as a key input into many policies and decisions through the pandemic: several government departments would eagerly anticipate the delivery of our report so that they could see the effect of their policy decisions unfold.

Two papers were published using data from the project.

What were some key insights?

I learned a huge amount from this project, but one insight seemed particularly key: that governments should probably do more living surveys to track public behaviour.

If our living survey was so useful for Covid-19, then I can see no reason why this method wouldn’t have benefits for other important social issues (e.g., racism, climate adaptation, adoption of artificial intelligence).

To better make the case for this, consider the following benefits which living surveys can provide to decision makers (compared to having no data about the public).

Aggregations of behaviours and attitudes allow decision makers to estimate approximate population-level adherence to key actions (e.g., recycling) and beliefs (e.g., that recycling is important to do).

Awareness of internal and unobservable behavioural drivers and barriers enable decision makers to look behind surface level actions to understand the root causes (e.g., differences in knowledge or belief) and how they differ between groups (e.g., adults in one geography and/or demographic versus another). This information is key for intervention design.

Forecasts for the future reduce uncertainty about the future. It is one thing to know that 50% of people are doing what you want, but that doesn’t say much about the trend. What if 75% of your audience plan to do the target behaviour next year? What if just 10% do? Advance knowledge gives you the benefit of advance preparation. This type of forecasting was key during the pandemic for problems such as transport planning.

Geodemographic audience segmentation is significantly better than having just geographic or demographic information, or neither. For instance, you might find that most littering is done by one demographic in three specific government areas. This allows you to be much more efficient in responding than if you have to address all audiences in all government areas. Plus, you reduce the risk of fatiguing and annoying compliant parties with irrelevant messaging.

Intervention tailoring is far superior to mass messaging. There is a reason why commercial marketers focus so much on segmentation: receptiveness to different messages varies hugely by audience. What works for Australian teens may not necessarily work for older immigrants who pay attention to different sources, hold different beliefs and face different barriers.

In summary

I have argued that the SCRUB survey (and similar work during the pandemic) showed that governments can benefit a lot from using ‘living surveys’ to better understand the role of public behaviour related to many important issues (e.g., recycling, mental health, or technology)

Right now governments who aren’t doing this are missing out on the benefits of knowing: i) what different demographics/geographies/groups do or think; ii) why people fail to act as hoped (i.e., it could be because they are unaware, unable, or unmotivated); iii) how people expect to think or act in the future; iv) who is best to target and; v) what to say to whom to get the best outcomes.

The solution may be relatively simple (at least in theory): Government should design, pilot and test living surveys in relevant contexts, then scale up those which provide cost-effective insights.

Acknowledgements

The above work was conducted with Alexander Saeri, Emily Grundy, Liam Smith, Morgan Tear, Micheal Noetel and BehaviourWorks Australia.

Done reading? Make sure to share your own thoughts about how living surveys can enhance governments' understanding of public behaviour by leaving a comment below ⬇️

(Image credit: Unsplash)