Carrying out research during lockdown: how to do remote surveys in rural communities

Mercy Corps and Practical Action’s report Avoiding a perfect storm: COVID-19 and floods in Nepal is based on findings from a remote, telephone, survey we conducted with 46 Community Disaster Management Committees (CDMC) which we work with as part of our programmes within the Zurich Flood Resilience Alliance. As we are sure others are having to change the way they collect information or data from communities we thought we’d share some of our reflections from this experience. 

Sharing objectives and choosing the right questions

It’s always important to make sure all involved in a project know and understand the objectives of it, this requires extra attention when multiple organisations are involved. By writing up and sharing the objectives of the survey and circulating this throughout the teams we made sure everyone had a shared understanding of the survey purpose and the information we were looking for.

To identify the right questions to ask, and ensure we had the foundational knowledge we needed, we reviewed a range of context analysis guidance documents and resources related to the sectors we wanted to learn more about, for example:

• Risk Communication and Community Engagement (RCCE) Action Plan Guidance COVID-19 Preparedness and Response (WHO)
• COVID-19: How to include marginalized and vulnerable people in risk communication and community engagement (WHO, IFRC, UNOCHA)
• Disaster Resilience Scorecard for Cities Public Health System Resilience (UNDRR)
• Interim Guidance on Scaling-up COVID-19 Outbreak in Readiness and Response Operations in Camps and Camp-like Settings ( IFRC, IOM, UNHCR and WHO)
• The Sphere standards and the Coronavirus response

We made sure to draft questions in a way and language that was accessible to the CDMC members we surveyed, using plain Nepali. We involved the Project Coordinator and Social Mobilisers (colleagues working closely within the community) for input as they have the best understanding of community context, and were the staff carrying out the surveying.

We tried to limit the number of questions in order not to take up too much of respondent’s time.

It is vital that those conducting the survey are comfortable with all questions and understand them well enough to explain if something isn’t clear to the respondent. In order to make sure that the staff was confident with the surveys they practiced among themselves, only when they felt comfortable did we go live with surveying the CDMCs.

Using appropriate technology

The CDMC members in the communities where we work all have access to phones, hence that’s the technology we chose to use. Online surveys sent via email were ruled out as not all respondents would have access to internet or devices that would be suitable.

Doing a survey using the telephone also improves the response rate. You’re more likely for someone to respond when you talk to them directly, are able to explain the purpose of the survey and how the respondent will benefit from participating, and answer any initial questions they might have. If you catch someone at a time that doesn’t suit them, or something demands their attention away half way through, you can easily agree a more suitable time and call them back.

Some of the surveys were carried out using ONA, an android based mobile data collection/survey platform. Where this was used the Social Mobiliser inputted responses directly into a form built in the app, saving time as responses didn’t need to be transcribed and retyped when shared digitally, instead mobile data could be used to share results with colleagues in Kathmandu.

Team work in the face of lockdown

Some colleagues preferred having the survey questions printed out and responses recorded using pen and paper. A challenge as most field staff do not have access to a laptop, only have internet access via their phones, and cannot access a printer during lockdown. Instead the Finance Officer who is still required to attend the office at times agreed to print the questions for us.

Still this left the challenge of getting the print outs to our colleagues on lockdown in communities we were not able to access. Local government however are able to access these communities to provide poor households with food rations and sanitation resources. As we have a good relationship with local government, they were happy to do us this favour as they were already visiting communities.

Collecting the data

Make sure you find a quiet space where you won’t get interrupted. This can be easier said than done at these times of working from home surrounded by family members, children, and domestic responsibility. Communicate with the people you live with what you’re doing and how long it is likely to take.

Before jumping into the questionnaire, introduce yourself, explain why you’re calling, and what the purpose of the survey is. Including how the respondent will benefit from participating. Our team created a template on this information to get the understanding and consent of the respondents – this is important to ensure coherence and clarity across the different surveyors.

Our experience was that CMDCs were very interested to contribute, the overlapping risks of COVID-19 and Monsoon season are very relevant to them and they were keen to share their knowledge and experience to reduce these risks. It also helped that we have a long standing relationship with these groups where we are not seen as extracting knowledge, but rather making their voices and concerns heard. The three Social Mobilisers Practical Action and four from Mercy Corps recruited to carry out the survey are colleagues who work closely, and are familiar, with the communities and CDMCs they were surveying.

In the first introductory call set out how long the survey is likely to take and agree a suitable time to call them back to do the survey. Our experience is that we had to call people two to three times for all questions to be answered, each call taking between three and eight minutes. The reasons for multiple, short calls range from individuals, perhaps domestic responsibilities required attention, mobile reception wasn’t ideal, or they did not have all the information asked for and needed to look this up to give an accurate answer.

The opportunity the phone survey gave for people to go away and look up information, confirm with someone else in the CDMC or local health volunteers, or instruct you to reach out to someone else who would have the right knowledge was a bonus which wouldn’t necessarily happen in a physical situation. We still made sure to check details against existing data such as censuses, to make sure there weren’t unexplained discrepancies.

Sharing the data using a range of tools

For colleagues using the ONA platform data sharing with the team in Kathmandu was very simple. For others the sheet with responses was photographed and sent via Facebook Messenger to the project coordinator who digitised the data using a laptop. If any details were unclear he would call the Social Mobiliser to clarify.

The data was then forwarded to the team in Kathmandu via email. The Mercy Corps team manually aggregated the data in using Excel, this was manageable but if doing a survey with more participants and questions it would be more efficient to create and use the same answer template throughout the project.

Good enough in a world that’s anything but normal

In an ideal world, we would have liked to do focus group discussions in person and in-depth interviews, as well as surveys, when collecting data for a report like Avoiding a perfect storm: COVID-19 and floods in Nepal. But this is not an ideal world, we cannot meet in person and the urgent need for this information for governments and communities to take action, the time constraints for community members, and technological barriers meant we had to adapt and make the most of the solutions available to us.

If you like us need to find a good enough approach to gaining the knowledge you need to have an impact we hope you find our experience and learnings useful.

Learn from the results

If you’re interested in what we found from doing these surveys read our report Avoiding a perfect storm: COVID-19 and floods in Nepal.