|
| << previous | next >> |
Journal description
All volumes of the corresponding journal
All issues of the corresponding volume
All articles of the corresponding issues
Details for article 2 of 17 found articles
| Title: |
A Comparison of Methods which Utilize the Integrated Mean Square error Criterion for Constructing Response Surface Designs |
| Author: |
Ott, L. Cornell, J. A. |
| Appeared in: |
Communications in statistics |
| Paging: | Volume 3 (1974) nr. 11 pages 1053-1068 |
| Year: |
1974 |
| Contents: |
In recent years many authors have utilized the integrated mean square error (designated by J) criterion for developing response surface estimation procedures and for constructing response surface designs. In particular, the integrated mean square error criterion has been used with the method of least squares, with the method of minimum bias estimation and more recently, with a generalized least squares estimator. This paper is concerned not with the J criterion, but rather with the unintegrated properties, the variance and the bias, of the approximating polynomials arising from these three procedures. It is shown that although the three procedures differ very little according to their values of J, real differences may exist in their respective unintegrated properties. It is these unintegrated properties of the approximating polynomial which affect the estimation of the response surface as well as the subsequent analysis. |
| Publisher: |
Taylor & Francis |
| Source file: |
Elektronische Wetenschappelijke Tijdschriften |
Details for article 2 of 17 found articles
| << previous | next >> |