Developing A Customized, Enumeration Area-Based Sampling Frame Tailored to a Specific Population Subgroup Using Geospatial Methods

Abstract A national sampling frame typically comprises a list of Primary Sampling Units (PSUs), such as enumeration areas derived from census data, which are commonly used in household surveys. Both national statistical offices and non-governmental organizations often rely on this framework when conducting surveys related to forced displacement. However, these frames are generally developed without considering the estimated number or geographic distribution of displaced populations. As a result, achieving the desired sample size becomes difficult and cost-intensive, as selected units frequently contain no individuals of interest. This study aimed to evaluate the potential of geospatial methodologies to develop a digital national sample frame tailored to a specific population subgroup or the general population, with the goal of ensuring applicability across diverse settings. For the first time, this work produced publicly accessible, digitized boundaries for urban and rural areas in Cameroon that are aligned with official administrative divisions and do not follow a grid-based system. According to our classification and estimated number from the ProGres database, 46 percent of refugees in Cameroon resided in rural areas, while 31 percent lived in camps and 23 percent in urban settings. The proposed geospatial approach offers a cost-effective alternative to traditional manual methods, particularly in data-scarce environments, and eliminates common geometric inconsistencies found in manual mapping efforts. All sampling units were nested within administrative boundaries, and in populated areas, their delineations aligned with observable ground features and respected major physical barriers. Importantly, including the refugee population in the customized national sampling frame was essential, as it enhanced the representativeness of refugees within it. This approach can be easily adapted to other countries. Notably, it was implemented in preparation for 2024’s Forced Displacement Survey in Cameroon, highlighting its practical application and relevance in real-world survey contexts.