Novel class of population mean estimators based on robust regression methods

Area of Science:

• Statistics
• Survey Methodology

Background:

• Accurate population mean estimation in surveys is crucial but challenged by outliers.
• Traditional estimators can be biased or inefficient when data contains atypical observations.

Purpose of the Study:

• To propose a novel class of estimators for population mean under simple random sampling (SRS).
• To enhance resistance to outliers while maintaining efficiency using robust regression methods.

Main Methods:

• Formulation of estimators using robust regression techniques (Hample-M, Huber-M, Tukey-M, Huber-MM, LTS, LMS).
• Theoretical examination of bias and mean square error (MSE).
• Validation through simulation studies and real survey data analysis.

Main Results:

• Proposed estimators demonstrated superior performance compared to existing robust methods.
• Achieved minimum relative mean square error (RMSE) and maximum relative efficiency (RE).

Conclusions:

• The novel estimators offer a robust alternative for survey data with outliers.
• Effective for practitioners dealing with contaminated sample survey data.