Experimental design, statistical strategy

Max Morris is a Professor in the Statistics Department and in the Department of Industrial and Manufacturing Systems Engineering at Iowa State University. His research interests include the development and application of experimental designs and strategies for computer simulations, problems involving spatial and dynamic systems, and factor screening experiments. 

Statistical Techniques Experimental Design Optimization Strategies Multivariate Classification Techniques Multivariate Calibration Techniques Expert Systems Multicriterla Decision Making Signal Processing... 

See also Chemometrics and Statistics Experimental Design Optimization Strategies. 

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. 

This chapter constitutes an attempt to demonstrate the utility of multivariate statistics in several stages of the scientific process. As a provocation, it is suggested that the multivariate approach (in experimental design, in data description and in data analysis) will always be more informative and make generalizations more valid than the univariate approach. Finally, the multivariate strategy can be really enjoyable, not the least for its capacity to reveal hidden treasures in data that in a univariate analysis look like a set of random numbers. 

Besides analyzing and correlating data by statistical means, the chemical engineer also uses statistics in the development of quality control to establish acceptable limits of process variables and in the design of laboratory, pilot plant, and process plant (evolutionary operation) experiments. In the latter application, statistical strategy in the design of experiments enables the engineer to set experimental variables at levels that will yield maximum information with a minimum amount of data. 

On the contrary, using the strategy suggested by statistical experimental design [16] permits us to gain information on how to reach the co-ordinates of the optimum, with the lowest possible number of experiments. The merit of this approach is twofold not only does it permit us to save a lot of resources by dramatically diminishing the number of required experi-... 

Cui F, Zhao L. (2012). Optimization of xylanase production from Penicillium sp.WX-Zl by a two-step statistical strategy Plackett-Burman and Box-Behnken experimental design. Int J Mol Sci, 13, 10630-10646. 

ANOVA is a useful and practical statistical tool in developing a model from scratch. ANOVA as a design of experiment tool has proven to aid in optimization studies (Sudharsan and Ng 2000 Meador et al. 2009), parametric and behavioral studies (Nguyen et al. 2009) and development of empirical models (Montgomery 2013). ANOVA uses different types of experimental design techniques and statisticians can deploy whichever strategy that is deemed adequate to their research. 

We should point out that the ability to fit equations of the form (5-18) or like (5-19), or of an even more complicated form, is predicated on having data from enough different mixtures to allow unambiguous identification of the parameters b. This requires proper data collection strategy. Much of the existing statistical research on the topic of mixture experiment design has to do with the question of wise allocation of experimental resources under the assumption that a particular type of equation is to be fit. 

Well-designed, appropriate experimental strategies will greatly enhance the effectiveness of a laboratory testing program. Statistically derived experimental strategies developed over recent decades provide the following benefits ... 

Many catalyst researchers are unaware of the potential benefits of statistical design of experiments. Others have had unfortunate experiences with so-called designed experiments because they underestimated the influence of experimental uncertainties on the reliability of the conclusions. In both cases familiarity with the fundamentals of statistical inference in the experimentation strategy is beneficial (4). Statistically derived strategies can certainly offer many important benefits, although they obviously cannot replace creativity or sound technical judgment. 

---