Akademik Veri Yönetim Sistemi
Journal of advanced research in natural and applied sciences (Online), cilt.11, sa.3, ss.224-236, 2025 (Hakemli Dergi)
This study introduces novel EWMA-based memory-type estimators for systematic sampling, addressing a gap in the literature where such estimators have not been developed for this sampling method. Systematic sampling is generally more efficient than simple random sampling due to its uniform coverage across the population. Building on existing research on memory-type estimators for various sampling techniques, this work presents three EWMA-based memory-type estimators: a ratio estimator, an exponential ratio estimator, and a regression estimator. Simulation studies using synthetic and real data reveal that the proposed estimators outperform existing methods in all conditions, with specific recommendations for using the ratio and regression estimators in different scenarios. The results suggest that the proposed methods, especially with a parameter value of ϑ = 0.3, are most effective for symmetric distributions.