This book addresses experimenters who have knowledge of classical experimental design methodology and expands their repertoire beyond hypothesis testing by providing statistical methods appropriate for selection, screening, and multiple comparisons. It concentrates on three types of procedures: selection procedures that use the "indifference-zone" approach, screening procedures using the "subset" approach, and multiple comparison procedures involving normal means. This is the first book, specifically designed for practitioners, to bring into focus many developments in the field previously covered only in university courses. It also presents new results on the comparison of procedures that have been obtained specifically for this volume.

This self-contained volume describes methods for designing experiments when the scientific objective is selection of best treatments, screening a set of treatments, and multiple comparisons among treatment means. The book emphasizes procedures appropriate in a variety of practical settings including those that require blocking and randomization restriction. It compares the relative merits of procedures when several different methods can be used in the same circumstances.

Providing practical guidance for experimenters in agriculture, engineering, medicine, and other empirical sciences, this book may also be used for a one-semester graduate course in selection methodology or to augment traditional courses in experimental design.

Design and Analysis of Experiments for Statistical Selection, Screening, and Multiple Comparisons:

• Shows how selection and screening can be applied to data that follow one of three important probability models—normal distribution, binomial distribution, and the multinomial distribution models
• Provides an extensive comparison of procedures, allowing experimenters to choose among competitors when several different procedures are feasible for a given application
• Gives an extensive set of tables of constants necessary to implement the procedures
• Supplements the tables of constants with listings of FORTRAN programs so that experimenters are not limited to those values covered by the tables
• Focuses on frequent formulations, while also providing references to Bayesian and other alternative developments in the Chapter Notes