Reliability of estimates in sociodemographic groups with small
Reliability of estimates in sociodemographic groups with small samples D. Buono & A. Bikauskaite 15 March 2017, NTTS, Brussels All expressed opinions are of the authors
Why interested in SAE? • European regional policies defining the variables of interest • Small sample sizes by some breakdowns • Reliable data breakdown demanded by policy makers and customers
Indicators of interest: ARPT
Estimation methods
tools and methods used • Empirical Bayes (EB) method based on the nested error model • Package: sae. R • Functions: direct eb. BHF pbmse. BHF • Hierarchical Bayes (HB) method based on the Fay. Herriot model • Package: hbsae. R • Functions: f. SAE. Area
Application: Target and data • Target: Calculate direct and indirect at-risk-of-poverty rate estimates by socio-demographic breakdowns • Data sources: Survey on Income and Living Conditions (EU -SILC) and Census data of some EU countries in 2011 • Sample: divided in 18 disjoint socio-demographic groups of small and large sizes • Auxiliary variables: unit level information on economic activity status and highest level of education attained
Application: Results
Learnings and future work • SAE techniques improve reliability of estimates • Solves the issue of no respondent above/below the threshold • Increase data availability • Further investigation is needed to assess the most appropriate estimator • Enlargement of number auxiliary variables • Harmonization of practices • Extension to additional countries and socio-demographic groups
- Slides: 8