Sadly, we cannot ask everyone in a country about their perspectives on risk due to time, accessibility and resource restrictions. So how did we prioritise?

Step 1: Selection of risk areas

We wanted to speak to people experiencing the full range of disasters that a country faces.

How?

A National Advisory Committee was tasked with mapping the vulnerabilities of each administrative area. To do this they used information from the INFORM website, national statistical websites, ReliefWeb, and NGOs’ recent vulnerability assessments. The top half of the administrative areas, those with the highest values, ranks or levels of vulnerability, were selected.

The group then used information from national hazard maps, the EmDat hazard lists, the Global Risk Data Platform and the INFORM platform to map which of these administrative areas were exposed to which hazards. These hazards could be natural, geological, environmental, conflict or technological in nature. Up to four administrative areas were then chosen that covered the range of hazards the country faces.

The number of risk areas selected depended on the diversity of risks within the country. For instance, if a country is very small, there might not be enough diversity in the vulnerability and hazards within the country to select 4 different areas. In this case, only 2 or 3 risk areas may have been chosen, or the whole country may have been viewed as one risk area. If on the other hand, the country is very big, with a high diversity of risks, then four risk areas were chosen to take into account this diversity.

Step 2: Selection of communities

Within these risk areas, we wanted to ensure that we were examining those communities with high levels of vulnerability, but also a balance of urban and rural contexts, and varying levels of government risk reduction engagement.

How?

The National Advisory Committees mapped out information for each community for each of the following criteria:

  • Be a balance of urban and rural
  • Have high levels of poverty
  • Be a balance of where there have been previous DRR interventions and where there have been less interventions
  • Be a balance of where there are active and inactive governments

Based on this mapping, 15 communities are selected equally across the risk areas.

Step 3: Selection of respondents

In choosing exactly who to survey we wanted to be both representative of the population as a whole, and ensure we captured the voices of those most marginalised. We also wanted to hear the perspectives of the full range of local actors: local CSOs, local government representatives and community members.

How?

To achieve these aims we conducted 4 surveys each with their own sampling methodology.

Random Household Survey: Interviews with individuals from households. The number of households to be sampled in a specific community is proportional to the population size of that specific community – approximately 10%.

In some communities, local CSOs or community leaders have a list or map of all the households in the community. In other communities, we asked the community to develop a map together. Each household was assigned a number, and a random number generator was used to select which households were sampled.

In order to overcome the potential bias of only certain types of individual being home during certain times of the day, alternating male and female family members were interviewed.

Community consultations: Focus group style discussions. In each of the 15 communities, five focus groups were organised. One each for women, men, children and youth, elderly and persons with disability. Each focus group contained at least 7 individuals.

Civil society organisation survey: 10 representatives from CSOs per community were interviewed. These were identified through previous connections with the partner organisations.

Local government survey: 10 representatives from local government units per community were selected using connections. These were identified through previous connections with the partner organisations.