Simple House Needs in Jember with Robust Small Area Estimation

  • Frida Murtinasari Program Magister Matematika FMIPA Universitas Jember
  • Alfian Futuhul Hadi
  • Dian Anggraeni Program Studi Matematika FMIPA Universitas Jember


SAE (Small Area Estimation) is often used by researchers, especially statisticians to estimate parameters of a subpopulation which has a small sample size. Empirical Best Linear Unbiased Prediction (EBLUP) is one of the indirect estimation methods in Small Area Estimation. The presence of outliers in the data can not guarantee that these methods yield precise predictions . Robust regression is one approach that is used in the model Small Area Estimation. Robust approach in estimating such a small area known as the Robust Small Area Estimation. Robust Small Area Estimation divided into several approaches. It calls Maximum Likelihood and M- Estimation. From the result, Robust Small Area Estimation with M-Estimation has the smallest RMSE than others. The value is 1473.7 (with outliers) and 1279.6 (without outlier). In addition the research also indicated that REBLUP with M-Estimation more robust to outliers. It causes the RMSE value with EBLUP has five times to be large with only one outlier are included in the data analysis. As for the REBLUP method is relatively more stable RMSE results.

Author Biography

Alfian Futuhul Hadi


How to Cite
MURTINASARI, Frida; HADI, Alfian Futuhul; ANGGRAENI, Dian. Simple House Needs in Jember with Robust Small Area Estimation. Jurnal ILMU DASAR, [S.l.], v. 18, n. 1, p. 1-8, feb. 2017. ISSN 2442-5613. Available at: <>. Date accessed: 19 july 2018. doi: