A New Method for Estimating Error Variances in Split-Plot Experiments  

Oleh:

Ikeda, Shu ; Matsuura, Shun ; Suzuki, Hideo

Jenis: Article from Proceeding
Dalam koleksi: Asian Network for Quality (ANQ) Congress 2011, Ho Chi Minh City, Vietnam, 27-30 September 2011, page 1-9.
Topik: Split-Plot Design; Response Surface Model; Generalized Least Squares; Restricted Maximum Likelihood
Fulltext: JP30_Ikeda_Fullpaper.pdf (158.12KB)

Isi artikel

In many industrial experiments, completely randomized designs are not practical in terms of time and cost, and split-plot designs are frequently used. For estimating response surface models in split-plot experiments, estimating the components of the covariance matrix of observations is often necessary. This study proposes a new method for estimating error variances to use Feasible Generalized Least Squares procedure. Our method is more accessible for practitioners than Restricted Maximum Likelihood procedure recommended in many previous studies. Some simulation studies are conducted to evaluate the performance of our method. The results of them indicate that our method performs well enough to be an alternative of the existing methods.

Opini Anda

Klik untuk menuliskan opini Anda tentang koleksi ini!

Kembali