Model-assisted estimators based on clustered coefficient linear regression models in small area estimation

Model-assisted estimators using sampling weights are employed in small area estimation problems to provide reliable estimators for small areas. Unlike traditional models that assume common regression coefficients for all areas, new estimators based on clustered coefficient linear regression models are proposed. These proposed estimators allow for heterogeneity of regression coefficients across different areas. Pairwise penalties are used to identify clusters and estimate parameters simultaneously. In the simulation study, we evaluate and compare the performances of proposed estimators with existing estimators. Additionally, we apply the new estimators to a dataset from the National Health and Nutrition Examination Survey.