Alive & Thrive Nigeria Phase 1 impact evaluation endline household survey

This survey is part of a cluster-randomized evaluation of the impact of the Alive & Thrive (A&T) Nigeria program on infant and young child feeding (IYCF) practices. Data were collected in 23 local government areas in Kaduna State and 16 local government areas in Lagos State using a population-based sampling strategy. Mothers were eligible if they had a biological child 0 to 23 months of age and were 18 years of age, regardless of their marital status, or 15 to 17 years of age and married. The purpose of the endline was to document key IYCF practices in the intervention and control areas after the A&T and to assess the impact of the intervention using difference-in-differences analysis. Stata MP, Version 16 Sample and sampling procedures: Two methods were used to draw a random sample of eligible mothers. The primary method was a form of gridded population sampling, called geosampling, which represents a multistage stratified clustered design. The geosampling method did not produce an adequate sample size in some subgroups of interest (e.g., eligible mothers with children 0 to 5 months), so a supplemental sample was designed and fielded using random walk. // Weighting: Respondents from the two sampling methods (geosampling; random walk) have been pooled for analysis. Due to the complex design, data users are advised to produce estimates and standard errors using the final analysis weight (W_FINAL). It is recommended that standard errors be produced assuming a two-stage “with replacement” design – this is simpler approximation of the actual complex multistage design. First-stage strata are defined by VARSTRAT, and first-stage clusters are defined by VARCLUST. Some sampling strata have been collapsed to ensure a minimum of two clusters per stratum for estimation of within-stratum variance. To calculate the final analysis weight (W_FINAL), the base weights were first calculated separately for each sampling method as the product of the inverse probabilities of selection for each stage of sampling.