Innovative memory-type calibration estimators for better survey accuracy in stratified sampling

Area of Science:

• Statistics
• Survey Methodology
• Data Analysis

Background:

• Calibration methods enhance parameter estimation by integrating diverse data sources.
• Memory-type statistics (EWMA, EEWMA, HEWMA) utilize historical and current data for population parameter estimation.

Purpose of the Study:

• To introduce novel ratio and product estimators within a calibration framework using memory-type statistics.
• To evaluate the performance of these new estimators against existing methods.

Main Methods:

• Development of new calibration-based ratio and product estimators incorporating EWMA, EEWMA, and HEWMA statistics.
• Conducting a simulation study to compute Mean Squared Error (MSE) and Relative Efficiency (RE).
• Validation through a real-world application.

Main Results:

• The proposed calibration-based memory-type estimators demonstrated superior performance.
• Consistently lower MSE and higher RE were observed compared to existing memory-type estimators.
• Graphical representations illustrated the robust behavior of the new estimators.

Conclusions:

• The novel calibration-based memory-type estimators offer significant improvements in accuracy and efficiency.
• These methods provide a more effective approach to population parameter estimation.
• The findings confirm the superiority of the proposed estimators over traditional techniques.