Designing Experiments and Analyzing Data: A Model Comparison Perspective

Rating:
Summary: This is a scholarly presentation of a difficult subject.
Review: I continue to select this text for a graduate course in advanced experimental design. The book is not easy, and that is one reason why I like it. It presents a thoughtful scholarship on controversial issues pertaining to the use of inferential statistics as a tool to assist the researcher in making decisions about the validity and strength of functional relationships present in the data. The book eliminates the naive overconfidence that sometimes results from cookbook applications of algorithms to the results of experiments. It shows that the process of discovery and data analysis is never complete and that the general linear model is an imperfect technique by which to discern orderliness in nature where intersubject variability is a given.

Rating:
Summary: A good book but a terrible textbook
Review: This is a practitioner's book, not a scolar's. I'm not a scolar -- never have been and never will be. I love this book. It's lucid, it's sensible and it's great statistics. It goes thoroughly into the logic of linear model ANOVAs yet the bulk of the exposition is in the simple English language -- it has no calculus and no eigenvectors and almost no matrix algebra. You need an acquaintance with elementary one-way ANOVA but no more background than that. The authors pace it carefully and are not afraid of a bit of repetition for the sake of clarity (something I always appreciate when I'm reading new technical material). It's one of the best $105 dollars I've ever spent. I got loads out of it all the way through and it's a big book. The explanation of multivariate repeated measures is worth the price alone.

Rating:
Summary: Fabulous book
Review: This is a practitioner's book, not a scolar's. I'm not a scolar -- never have been and never will be. I love this book. It's lucid, it's sensible and it's great statistics. It goes thoroughly into the logic of linear model ANOVAs yet the bulk of the exposition is in the simple English language -- it has no calculus and no eigenvectors and almost no matrix algebra. You need an acquaintance with elementary one-way ANOVA but no more background than that. The authors pace it carefully and are not afraid of a bit of repetition for the sake of clarity (something I always appreciate when I'm reading new technical material). It's one of the best $105 dollars I've ever spent. I got loads out of it all the way through and it's a big book. The explanation of multivariate repeated measures is worth the price alone.

Rating:
Summary: A good book but a terrible textbook
Review: Without any doubt, this is a comprehensive book in this field. However, because the authors tried to demonstrate their superior english writting abilities, they made the whole text hard to understand for students. From my point of view, as a new professor, I don't believe that recommending this book is appropriate.