Building Resilience and Adaptation to Climate Extremes and Disasters 2019

Our data collection was carried out by first recruiting and training the interviewers, and then mobilising them in the field for a period of 6 days. We then carried out a remote control based on the data collected daily and the GPS coordinates collected automatically.

Training and deployment of interviewers

Firstly, for the recruitment of interviewers, we identified a group of 12 interviewers with the following characteristics:

• Each of them had at least 7 years of survey experience;
• Each spoke the most common Fulani dialect in the area correctly;
• Each was almost 100% ethnic Fulani, so they were able to establish a relationship of trust with the respondent;
• Each had already carried out at least one survey in a pastoral environment, during which questions on the size of the herd were asked.

After recruiting the interviewers, we trained them over a two-day period. During these two days, we shared with them the objective of the study, the principle of index-based agricultural insurance, its methodology in terms of contingent valuation of willingness to pay and the identification of the person to be surveyed in the household after having identified the household to be surveyed. The questionnaire was converted into a data collection application or electronic questionnaire. This application was tested on the last day of training through simulated survey exercises that also allowed for the oral translation of all questions into Fulani. This exercise enabled the understanding of the questions to be harmonised between the interviewers to reduce the risk of discrepancies in the explanations given to the farmers.

The stages of the training were as follows:

• Detailed explanation of the methodological aspects;
• Presentation of the objectives of the survey;
• Silent reading of the paper questionnaire "Survey Yourself": in this exercise the agents, while reading the questionnaire, play the role of both the interviewer and the respondent. This pedagogical exercise allows the interviewers to ask themselves different questions even before reading the questionnaire in plenary;
• Reading the questionnaire, question by question;
• Training of interviewers on smartphones and the collection application;
• Translation of the questionnaire into the local language (Pulaar);
• Paired survey: during this exercise, the interviewers will be paired up; one agent plays the role of the interviewer and the other plays the role of the respondent. Then the roles are reversed.

Coordination of surveys and data control

After the training, the teams were divided into groups of four people. In each team, there were three interviewers and a supervisor responsible for coordinating the team's actions and reporting regularly on the field situation to the study coordinator and statistician who are based in Dakar. As 15 communes were chosen for the surveys, they were divided into three, with five communes to be surveyed per team. To give each team freedom of movement, a vehicle with a driver was provided for each team. Mission orders and letters of introduction to be presented to the local authorities were also made available to them to facilitate their first contact with the local population and to have elements of proof of their good faith.
The team supervisor was responsible for reporting to the statistician in charge of the survey and the survey coordinator any difficulties or issues identified in the execution of the survey by their team. The team's statistician ensured that the data were reported daily and, on the basis of this, debriefed the supervisors on the quality of the data collected. The data collected through the tablets was uploaded via the Internet to a web platform set up for this purpose. The technology we used for mobile data collection is called ODK Collect. It is a suite of tools that allows data to be collected using mobile devices such as smartphones and/or tablets (running Android) and to submit the same data to an online server.