STATFRAM Functions
The Statistical Framework team (STATFRAM) developed a series of functions listed below to carry out analyses relevant to inform policymaking decisions. Each function uses either survey or routine data, and some allow users to bring their own data which allows gives the flexibility necessary to address multiple types of NEP questions. All of these functions are programmed in R programming language and are hosted on Stats Report.
§ getEstimates – calculates national, regional or district‐level estimates and standard errors within selected strata of maternal and child health indicators. This was used to do district-level estimation of coverage for indicators related to the top 3 causes of death in Mozambique.
§ getContrast – calculates differences and their standard errors between pairs of 2 or more geographic locations and/or strata. This was created for NEP-Tanzania to compare indicator rates within the same year.
§ getTrend – calculates the differences with standard errors across between rate estimates at two time points for geographic units or strata. This was created for NEP-Tanzania to compare indicator rates between different years.
§ doReg – performs a regression analysis with DHS data. Generalized linear models are used to allow binary, count, survival and continuous outcomes. This was used for NEP-Mali’s Cycle 2 question to identify what are some factors that might explain the differences between the districts of Mopti and those of other regions for two representative indicators.
§ visualizeOutliers – creates a boxplot in order to visualize outliers among predictors. This was used to assess the distribution and data quality of the routine regression covariate data used in NEP-Mali's Cycle 2 question.