Analysis of Messy Data: Designed Experiments v. 1

This practical tool for statisticians offers techniques and methods for effectively analyzing non-standard or messy data sets that arise from experimental design situations. The volume focuses on the analysis of variance techniques, covering the more basic ones in early chapters, including one-and two-way analyses of variance and multiple-comparison procedures. It also provides a unique approach to experimental design, which emphasizes the distinction between design structure and the structure of treatments. The middle portion of the book deals with unbalanced data in two-way structures. Here, the book describes and uses different linear models, the so-called means model, and the effects model, with some treatment of higher-order structures. The book then moves on to random and mixed models, stressing the estimation of, and inference about, variance components. The final chapters focus on more complex structures, including designs with several sizes of experimental units, such as split-plot designs and repeated-measure designs.
Throughout, the book introduces each topic with several examples, follows up with a more theoretical discussion, and concludes with a case study using actual data. Its overriding emphasis on practical implementation extends to computers, with several available statistical packages, including SAS, BMD and SPSS.
 
This practical tool for statisticians offers techniques and methods for effectively analyzing non-standard or messy data sets that arise from experimental design situations. The volume focuses on the analysis of variance techniques, covering the more basic ones in early chapters, including one-and two-way analyses of variance and multiple-comparison procedures. It also provides a unique approach to experimental design, which emphasizes the distinction between design structure and the structure of treatments. The middle portion of the book deals with unbalanced data in two-way structures. Here, the book describes and uses different linear models, the so-called means model, and the effects model, with some treatment of higher-order structures. The book then moves on to random and mixed models, stressing the estimation of, and inference about, variance components. The final chapters focus on more complex structures, including designs with several sizes of experimental units, such as split-plot designs and repeated-measure designs.
Throughout, the book introduces each topic with several examples, follows up with a more theoretical discussion, and concludes with a case study using actual data. Its overriding emphasis on practical implementation extends to computers, with several available statistical packages, including SAS, BMD and SPSS.