Written byKrittika Gorur, Pooja Gupta, Mark Milrine and Mareike Schomerus (Busara Center for Behavioral Economics) and originally published on the Elrha website.

Illustration of coronavirus. Credit: Center for Disease Control and Prevention (CDC)

One characteristic of a crisis is that what you need most can be most difficult to get.

When COVID-19 and rumours about the virus were spreading across the world, what was needed most was information: information on the disease. Information on how to prevent it. Information about why misinformation spread as quickly as — if not faster than — the virus itself. The latter caught our attention: what, exactly, made people believe misinformation?

For many people, COVID was not the only crisis they were facing. COVID was, for them, only part of a multi-layered, more complex emergency situation. People who had been displaced by conflict, violence or other disasters found themselves struggling to deal with the pandemic in already challenging circumstances. Rumour mills were busily spinning in refugee camps in central and east Africa, producing word on the street on possible cures or spreading wrong information on social distancing and mask wearing. Misinformation was rife in this multi-crisis of COVID and displacement.

Acutely aware of this, we wanted to conduct research on how misinformation — which is defined as wrong or damaging information that is often shared unknowingly — spreads in refugee communities, and to identify how behavioral science can help counter this phenomenon. The only possible way to conduct this research in such an acute moment was through a partnership with a humanitarian organisation.

What humanitarian organizations working in displacement settings most needed at that point was to urgently understand how to design information campaigns that would convince people to adhere to public health advice and dismiss misinformation.

What Busara as a research organisation most needed in order to give substantiated advice on how to counter misinformation was to implement sophisticated research with complex data collection.

The challenge of delivering such a complex research project with slightly different organisational needs was intimidating. Health education is difficult enough under normal circumstances. Misinformation in the midst of a pandemic, spreading amongst people who have experienced displacement, fear and violence can make the situation significantly worse for everyone. We quickly got to work, designing rigorous research with behavioral methods.

It turns out that in an acute crisis, research and humanitarian needs are very challenging to combine. While we were able to conduct and learn from a pilot study, we never managed to fully implement our ambitious research design (although we have produced background work, a research agenda and insights into behavioral mechanisms in misinformation). Yet, our biggest lessons sometimes come from what we fail to achieve.

Here, we reflect on what we learned on attempting behavioral research in humanitarian settings during a crisis and suggest what we would do differently next time.

In a crisis, be partners before partnering

Rapid COVID-19 research also created rapid partnerships, with little time to understand different organisation’s ways of working. While rapid COVID behavioral research was alluring and pioneering, partnerships between organizations doing very different work need time to grow. Researchers and humanitarians could swiftly agree on a research need, but barriers to delivery crystallized only over time: staff buy-in, implementation of data collection, and a lack of short-term benefits materialising from the project which might have made it attractive for staff to spend their time supporting it over the long-term. Because the partnerships we formed rapidly were new, we did not have a clear idea of organizational needs and underestimated logistics, time, and resource commitments.

In established partnerships, it is easier to know what organizations are committing to. Ways of developing proposals are more rehearsed and it might be easier to include multiple perspectives early on to muster sufficient buy-in. Discussing risk mitigation strategies before submitting a proposal can go a long way towards managing any conflicts that may arise. This is especially true if researchers come in to work in communities where a humanitarian agency has often spent years building trust and nurturing relationships. Researchers — even if they come as partners of the humanitarian organization — can unwittingly upset this delicate trust and it is crucial to identify before committing to research what might be perceived as disruptive, and discuss protocols to monitor researchers’ interactions with local communities early on.

It’s tough, maybe impossible, to do partnership ground work in an acute crisis. Working through established partnerships in such situations is likely a better choice.

Never forget that there is a crisis

When we started designing the research, COVID-19 was new and unpredictable. Yet, because we wanted to conduct rapid research, we only took a snapshot assessment of the situation and planned all our engagement from there. We assumed how much time we could expect from field officers, how technology would support our data collection, and that there would not be another version of the crisis. But we forgot that what makes a crisis a crisis is unpredictability. We did not adequately plan for a rapidly deteriorating COVID situation that stretched and exhausted staff (with crucial staff leaving and being replaced by others new to the research project).

Although our plans had factored in the low levels of mobile phone usage in the study settings, based on best available information about numbers of phone users, once the study was underway it became clear the numbers were even lower than our partners realised and we had expected. While we conducted a standard risk assessment before the project started, the reality of a military intervention disrupting the security situation was not something we considered seriously enough: and because of the rapid nature of this study, probably did not invest enough time in planning for.

Working in a crisis moment in conflict-prone environments means that the accompanying stressors and uncertainties are unlikely to be accommodating. There is no short cut for nuanced risk assessments and scenario planning to support adaptation.

Developing data use agreements takes time

Crisis research needs to be swift. Our idea was to team up to allow for rapid research implementation by reaching populations through established humanitarian channels. But for humanitarian agencies, sharing contact data of refugees with third-party researchers poses a significant ethical and legal dilemma — even if researchers only use the contact data to ask for consent to participate in a study. In what was supposed to be a pioneering study, it turned out that we did not fully consider the details of these ethical dilemmas.

The discussions that ensued are necessary — but they don’t gel with the notion of rapid research. Again, with an established partnership, it is likely that a data-sharing and management agreement could have been worked out in advance as part of a memorandum of understanding or established partnership agreement.

Money talks

When staff changes, as they do frequently in both research and humanitarian organisations, budgets remain–and the numbers can help keep the conversation going. Budget planning is often awkward (it means disclosing how each organisation works and accountability requirements can be complex). As a result, it is often done with top-level figures only, missing the opportunity to make it a process to secure commitment from partners.

When there is staff turnover throughout the life of a partnership, jointly agreed budgets can act as a reminder of commitments made. If a project is taken over by a new leader (who may not be as enthusiastic about it or might be under pressure from other projects), having financial resources firmly allocated for each task can help persuade staff to continue a partnership. Our budget was primarily driven by research needs, prioritising partners over others, rather than by the idea of it being a tool for a long-haul partnership.

Find ways of positively engaging field staff

The benefits of research often take some time to materialise and thus are not obvious for humanitarian field staff. All research seems to do, in the moment, is to add workload to already overstretched people dealing with different scenario and contingency plans. For a humanitarian/research partnership to work, staff at all levels in all organisations need to be able to commit to spending time with each other to understand the value of the research for their work. If this time cannot be found, research is likely to feel mismatched and burdensome, making it much more difficult to implement.

Regular project updates to field staff and cocreation session on goals for a research project are necessary. But even with those in place, it is imperative to highlight continuously how much time commitment this kind of undertaking really requires — particularly if humanitarian staff is not to act simply as an enabler of a research project, but an active and substantive contributor, too.

Our failed project taught us a lot of lessons

In the heady days of wanting to do our part to contribute to learning on COVID misinformation in refugee settings, the main driver of our project was to make sure that we were making a contribution to rapid COVID-19 research. But combining complex research with equally complex humanitarian settings does not just happen. To make it work requires taking all of the above into account.

For us as researchers, we realised that moments of crisis make it even harder to be honest about our own limits, about organisational needs that cannot be short-circuited, and about how to best care for communities who are supposed to benefit from research.

Honest reflection after a failed research project can be painful. But it is necessary, because the difficult task has not changed: In order to address a crisis, what we need most are rigorous research findings to inform on the ongoing crisis. What we need most is also the most difficult to get — including keeping a cool head when planning complex research and, in future, returning to the lessons learned that we shared here and commit to what we now know we need to do differently.