Title: Split-plot experiments with time trends in easy-to-change factors

Authors: Mubashir Siddiqui, Kai Yang

Addresses: Department of Industrial and Manufacturing Engineering, NED University of Engineering and Technology, University Road, Karachi, 75270, Pakistan. ' Department of Industrial and Manufacturing Engineering, Wayne State University, Detroit, MI 48201, USA.

Abstract: Split-plot experiments involve two types of factors, hard-to-change factors (HTC) and easy-to-change factors (ETC). In regular split-plot experiments, the experimental sequence of ETC factor levels should be randomised. However in many practical applications, the run sequence of ETC factors can only be changed in one direction. This restriction in the randomisation of ETC factors will create many serially correlated response observation subsequences, and regular analysis procedure will no longer be valid due to non independence of experimental residuals. In this paper, an integrated factorial and time series model is developed to analyse this kind of split-plot experiment with time trend in ETC factors.

Keywords: split-plot design; randomisation; time series; hard-to-change factors; HTC; easy-to-change factors; ETC; split-plot experiments.

DOI: 10.1504/IJQET.2010.034614

International Journal of Quality Engineering and Technology, 2010 Vol.1 No.3, pp.317 - 338

Published online: 13 Aug 2010 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article