Minimum covariance determinant-based estimators for estimating population mean using two auxiliary variables

Zaman T., Islamoglu E., Bulut H.

Maejo International Journal of Science and Technology, vol.20, no.1, pp.108-123, 2026 (SCI-Expanded, Scopus)

  • Publication Type: Article / Article
  • Volume: 20 Issue: 1
  • Publication Date: 2026
  • Journal Name: Maejo International Journal of Science and Technology
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Aerospace Database, Directory of Open Access Journals, Academic Search Ultimate (EBSCO), East & South Asia Database (ProQuest), Natural Science Collection (ProQuest), Biological Science Database (ProQuest), Biomedical Reference Collection: Corporate Edition (EBSCO), Earth, Atmospheric, & Aquatic Science Collection (ProQuest), Materials Science & Engineering Collection (ProQuest), Technology Collection (ProQuest)
  • Page Numbers: pp.108-123
  • Keywords: mean sqaure error, minimum covariance determinant estimate, ratio estimators, robust regression
  • Ondokuz Mayıs University Affiliated: Yes

Abstract

This study addresses the challenge of outliers in sample surveys, a common issue that can distort results and lead to inaccurate conclusions. To tackle this problem, researchers have developed various robust regression techniques including least trimmed squares, least median of squares, least absolute deviations, Huber’s, Hample’s and Tukey’s maximum likelihood estimators, and Huber’s maximum likelihood estimator for location and scale estimators. We propose a new approach: minimum covariance determinant-based ratio estimators for estimating the population mean in sample surveys. To assess the performance of the proposed minimum covariance determinant-based estimators, we derive their mean square error expressions and determine the conditions under which they surpass existing estimators in efficiency. To validate our theoretical findings, we conduct a numerical illustration using real-world data and perform simulations with R-Programme software. The results demonstrate that our method effectively handles outliers and improves the overall accuracy of survey-based estimations, making it a valuable tool for researchers working with sample survey data.