Response Surface Methodology: 1966-1988

Abstract

The authors intent is to provide a review of important developments in response surface methodology, with the definition being confined to that of a collection of tools in design or data analysis that enhance the exploration of a region of design variables in one or more responses. They emphasize developments in the statistics literature that have appeared since 1966 but make reference to some pre-1966 work where historical perspective is necessary. In terms of subject matter applications, the authors do not confine themselves to applications in the chemical and processing fields. They feel that one of the more important items to report is the breadth of application of RSM that has evolved in the last 20 years. Applications have expanded to areas such as operations research, nuclear energy, defense systems research, cancer chemotherapy, and many others. This document reviews the spectrum from theoretical developments to practical notions of current software availability in both the design and analysis aspects of RSM. Design and analysis will be treated separately though the user must treat the two together in his total analysis. The original intent by Box and company was to highlight a strategy-- both experimental and analytic. Yet precious little research employs a simultaneous consideration of both of these aspects. Finally, suggestions regarding the important areas for future research and future impact of RSM are offered. Keywords: experimental design; design variables.