Journal of Japan Society of Civil Engineers, Ser. D3 (Infrastructure Planning and Management)
Online ISSN : 2185-6540
ISSN-L : 2185-6540
Infrastructure Planning and Management Vol.28 (Special Issue)
EFFICIENT SP EXPERIMENT DESIGN BASED ON RP ATTRIBUTE VALUES: FOCUSING ON ERROR COMPONENTS BETWEEN RP AND SP
Nobuhiro SANKOToshiyuki YAMAMOTO
Author information
JOURNAL FREE ACCESS

2011 Volume 67 Issue 5 Pages 67_I_427-67_I_434

Details
Abstract

RP/SP model has been widely used, but an efficient SP experiment design for RP/SP model has not been examined very much. This study aims at investigating estimation efficiency of RP/SP models considering both the design of the SP experiment and the assumptions made with respect to error structures. The attribute levels of the SP survey are sometimes designed by considering their differences from those of the RP survey. In the RP/SP model, assumptions are made about error structures between the RP and SP models. However, it is not known how the SP design and the assumptions are related from the viewpoint of estimation efficiency of RP/SP models. This study analyses the relationship using simulation technique. Both a true model, where the model structure is the same as the one assumed in the data generation process, and a general model, which includes the true model as a special case, are estimated. Results suggest that the larger the variance of error components, the larger the difference in attribute values between the RP and SP required to obtain better efficiency, and that better efficiency of general models is obateind when the attribute values of the SP is close to those of the RP.

Content from these authors
© 2011 Japan Society of Civil Engineers
Previous article Next article
feedback
Top