Abstract
In split-plot experiments, estimation of unknown parameters by generalized least squares (GLS), as opposed to ordinary least squares (OLS), is required, owing to the existence of whole- and subplot errors. However, estimating the error variances is often necessary for GLS. Restricted maximum likelihood (REML) is an established method for estimating the error variances, and its benefits have been highlighted in many previous studies. This article proposes a new two-step residual-based approach for estimating error variances. Results of numerical simulations indicate that the proposed method performs sufficiently well to be considered as a suitable alternative to REML.
| Original language | English |
|---|---|
| Pages (from-to) | 342-358 |
| Number of pages | 17 |
| Journal | Communications in Statistics: Simulation and Computation |
| Volume | 43 |
| Issue number | 2 |
| DOIs | |
| Publication status | Published - 2014 Jan 1 |
Fingerprint
Keywords
- Generalized least squares
- Response Surface Methodology
- Restricted maximum likelihood
- Split-plot experiment
ASJC Scopus subject areas
- Modelling and Simulation
- Statistics and Probability
Cite this
Two-step residual-based estimation of error variances for generalized least squares in split-plot experiments. / Ikeda, Shu; Matsuura, Shun; Suzuki, Hideo.
In: Communications in Statistics: Simulation and Computation, Vol. 43, No. 2, 01.01.2014, p. 342-358.Research output: Contribution to journal › Article
}
TY - JOUR
T1 - Two-step residual-based estimation of error variances for generalized least squares in split-plot experiments
AU - Ikeda, Shu
AU - Matsuura, Shun
AU - Suzuki, Hideo
PY - 2014/1/1
Y1 - 2014/1/1
N2 - In split-plot experiments, estimation of unknown parameters by generalized least squares (GLS), as opposed to ordinary least squares (OLS), is required, owing to the existence of whole- and subplot errors. However, estimating the error variances is often necessary for GLS. Restricted maximum likelihood (REML) is an established method for estimating the error variances, and its benefits have been highlighted in many previous studies. This article proposes a new two-step residual-based approach for estimating error variances. Results of numerical simulations indicate that the proposed method performs sufficiently well to be considered as a suitable alternative to REML.
AB - In split-plot experiments, estimation of unknown parameters by generalized least squares (GLS), as opposed to ordinary least squares (OLS), is required, owing to the existence of whole- and subplot errors. However, estimating the error variances is often necessary for GLS. Restricted maximum likelihood (REML) is an established method for estimating the error variances, and its benefits have been highlighted in many previous studies. This article proposes a new two-step residual-based approach for estimating error variances. Results of numerical simulations indicate that the proposed method performs sufficiently well to be considered as a suitable alternative to REML.
KW - Generalized least squares
KW - Response Surface Methodology
KW - Restricted maximum likelihood
KW - Split-plot experiment
UR - http://www.scopus.com/inward/record.url?scp=84961386473&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84961386473&partnerID=8YFLogxK
U2 - 10.1080/03610918.2012.703280
DO - 10.1080/03610918.2012.703280
M3 - Article
AN - SCOPUS:84961386473
VL - 43
SP - 342
EP - 358
JO - Communications in Statistics Part B: Simulation and Computation
JF - Communications in Statistics Part B: Simulation and Computation
SN - 0361-0918
IS - 2
ER -