Title: Holliday's engine mapping experiment revisited: a design problem with 2-level repeated measures data

Authors: Harvey Goldstein, Toby Lewis

Addresses: University of Bristol, Bristol, UK. ' Centre for Statistics, University of East Anglia, Norwich, UK

Abstract: Only a limited amount of work has been done on issues of optimum experimental design in hierarchically structured data, in spite of the importance of design efficiency. We illustrate these issues by a fresh analysis of Holliday|s well-known engine mapping experiment. This experiment investigated the relation of automobile engine performance, as measured by engine torque vs. spark advance, to adjustable engine parameters or factors, viz. engine speed, load, and air/fuel ratio. We show how an optimum experiment can be designed. Our analysis suggests a general procedure of wide applicability.

Keywords: engine mapping experiments; optimum design; multilevel models; repeated measures; engine performance; experimental design.

DOI: 10.1504/IJVD.2006.009069

International Journal of Vehicle Design, 2006 Vol.40 No.4, pp.317 - 326

Published online: 22 Feb 2006 *

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