Abstract

The study evaluates the performance of a proposed methodology to synthesize regression coefficient estimates across different time periods. The method utilizes independently collected data sets and adopts the Multiple Imputation by Chained Equations (MICE) approach. Results demonstrated the feasibility of the methodology through empirical application using the bootstrap technique, which was implemented to assess accuracy and precision. Accuracy and precision were measured in terms of computed bias and bootstrap standard errors, respectively. Findings indicated that the regression coefficient estimates from the synthesized poverty model closely matched those obtained from the bootstrap technique, supported by negligible bias values. This suggests that the proposed methodology provides accurate estimates. Additionally, the low bootstrap standard errors indicate that the methodology yields precise estimates.