Statistical design overview

The strategy for robust design experiments that will be considered in Section 2.3 is based on the statistical techniques associated with response surface methodology. This section will give an overview of response surface methodology, presenting some of the more common experimental designs that have been developed in this area. 

In a robustness test the following steps can be identified (a) identification of the variables to be tested, (b) definition of the different levels for the variables, (c) selection of the experimental design, (d) definition of the experimental protocol, (e) definition of the responses to be determined, (f) execution of the experiments and determination of the responses of the method, (g) calculation of effects, (h) statistical and/or graphical analysis of the effects, and (i) drawing chemically relevant conclusions from the statistical analysis and, if necessary, taking measures to improve the performance of the method. A general overview of robustness testing can be found in [35). 

The first order of business in the study of a new reaction in the context of process research and development is to measure reaction rates, establish approximate reaction orders for empirical power-law rate equations, and obtain values of their apparent rate coefficients. This chapter presents a brief and critical overview of laboratory equipment, design of kinetic experiments, and evaluation of their results, including an overview of statistical methods. It is intended as a tour guide for the practical chemist or engineer. More complete and detailed descriptions can be found in standard texts on reaction engineering and kinetics [G1-G8],... 

Topics which arc covered in this overview include the role of the computational chemist in the design of new pharmaceutical and agricultural agents, along with descriptions of current trends in the field. Particular emphasis is placed on minicomputer graphics workstations, 3D stractural databases, and the merging of statistical and molecular modeling techniques to predict activity and mechanism. 

Part 1 of this section on analysis gives a survey of some of the more frequently used elementary analysis procedures along with a brief review of experimental design. The basic purpose of this section is to give a simple overview of the statistical concepts used in the application of these analysis techniques. Since computer programs are readily available for a more comprehensive analysis, a brief review of typical software analysis algorithms is also presented in Part 2. 

The first part of diis chapter has presented what we call "paper and pencil" tips regarding a variety of issues that can be considered when developing surveys and tests. One step in survey or test design is to compute a reliability statistic, such as Cronbach alpha. Typically, researchers collect data with their instrument and commence their statistical analysis. However, researchers should include a second analysis step that utilizes psychometric theory to guide die development of surveys and tests, to provide techniques Aat allow die functioning of surveys and tests to be monitored and improved, and to prepare data for statistical calculations. In the second part of the chapter, we provide an overview of one psychometric technique (the Rasch model) diat allows researchers to easily carryout this important second analysis step. 

---