Research will be conducted in three countries: Bangladesh, Sri Lanka and Zimbabwe and the independent but unrecognised territory of Somaliland. A household survey approach comprising of a baseline survey and two follow-up panel surveys is a central part of the research carried out in these locations. The baseline survey will collect representative information on each neighbourhood, concentrating on access to the five infrastructure grids (water, sanitation, energy, transport and communications). Panel surveys will help to trace how these patterns change through time. The second panel survey will also be part of the evaluation of the practical intervention.  

Six neighbourhoods will be identified in each country. Each neighbourhood will comprise between 1000 and 5000 households. These six neighbourhoods will consist of: two neighbourhoods involved in ‘The Unknown City’ project, to ensure continuity with previous research; two neighbourhoods from elsewhere in the capital city region to allow comparison within the city; and two neighbourhoods from elsewhere in the country. The purpose will be to collect a survey dataset essential for intra- and cross-country comparisons.  

Focus groups will inform the design of the initial survey instrument, ensuring that questions are as comprehensible as possible and are relevant to the local context with regard to services or infrastructure. The survey instrument will comprise a core section that will be used across all four countries with short country- and location-specific additions that will reflect the concerns of local stakeholders. The instrument was designed jointly by UK researchers and international partners in English then translated by international partners into four other languages. The survey will be conducted using tablets and the open-source Kobo Toolkit software with multi-language capability.  

Survey questions will be designed to get as accurate a picture as possible regarding access and use of grid infrastructure: the patterns of inclusion and exclusion in relation to the five services; the significance of gender, age, education, access to capital, etc. in ensuring access; alternative services and their value; advantages of inclusion; factors prohibiting access; and the implications of failure to comply with access rules. Questions shared across all countries will have standard responses; this maximises the potential for using the survey results for intra- and cross-country comparison.   

The baseline survey will cover between 10% and 20% of each neighbourhood, i.e. between 2000 and 3000 households in each city. The sample will be randomly selected. For the panel surveys, 10% of households already surveyed will be resurveyed in year 2 and the same 10% of households will be resurveyed again in year 3.