Designing Experiments and Analyzing Data: A Model Comparison Perspective

Designing Experiments and Analyzing Data: A Model Comparison Perspective

Designing Experiments and Analyzing Data: A Model Comparison Perspective

Designing Experiments and Analyzing Data: A Model Comparison Perspective


Through this book's unique model comparison approach, students and researchers are introduced to a set of fundamental principles for analyzing data. After seeing how these principles can be applied in simple designs, students are shown how these same principles also apply in more complicated designs.

Drs. Maxwell and Delaney believe that the model comparison approach better prepares students to understand the logic behind a general strategy of data analysis appropriate for various designs; and builds a stronger foundation, which allows for the introduction of more complex topics omitted from other books.

Several learning tools further strengthen the reader's understanding:

• flowcharts assist in choosing the most appropriate technique;

• an equation cross-referencing system aids in locating the initial, detailed definition and numerous summary equation tables assist readers in understanding differences between different methods for analyzing their data;

• examples based on actual research in a variety of behavioral sciences help students see the applications of the material;

• numerous exercises help develop a deeper understanding of the subject. Detailed solutions are provided for some of the exercises and
• realistic data sets allow the reader to see an analysis of data from each design in its entirety.

Updated throughout, the second edition features:

• significantly increased attention to measures of effects, including confidence intervals, strength of association, and effect size estimation for complex and simple designs;

• an increased use of statistical packages and the graphical presentation of data;

• new chapters (15 & 16) on multilevel models;

• the current controversies regarding statistical reasoning, such as the latest debates on hypothesis testing (ch. 2);

• a new preview of the experimental designs covered in the book (ch. 2);

• a CD with SPSS and SAS data sets for many of the text exercises, as well as tutorials reviewing basic statistics and regression; and

• a Web site containing examples of SPSS and SAS syntax for analyzing many of the text exercises.

Appropriate for advanced courses on experimental design or analysis, applied statistics, or analysis of variance taught in departments of psychology, education, statistics, business, and other social sciences, the book is also ideal for practicing researchers in these disciplines. A prerequisite of undergraduate statistics is assumed. An Instructor's Solutions Manual is available to those who adopt the book for classroom use.


This book is written to serve as either a textbook or a reference book on designing experiments and analyzing experimental data. Our particular concern is with the methodology appropriate in the behavioral sciences but the methods introduced can be applied in a variety of areas of scientific research. The book is centered around the view of data analysis as involving a comparison of models. We believe that this model comparison perspective offers significant advantages over the traditional variance partitioning approach usually used to teach analysis of variance. Instead of approaching each experimental design in terms of its own unique set of computational formulas, the model comparison approach allows us to introduce a few basic formulas that can be applied with the same underlying logic to every experimental design. This establishes an integrative theme that highlights how various designs and analyses are related to one another. The model comparison approach also allows us to cover topics that are often omitted in experimental design texts. For example, we are able to introduce the multivariate approach to repeated measures as a straightforward generalization of the approach used for between-subjects designs. Similarly, the analysis of nonorthogonal designs (designs with unequal cell sizes) fits nicely into our approach. Further, not only is the presentation of the standard analysis of covariance facilitated by the model comparison perspective, but we are also able to consider models that allow for heterogeneity of regression across conditions. In fact, the underlying logic can be applied directly to even more complex methodologies such as hierarchical linear modeling (discussed in this edition) and structural equation modeling. Thus, our approach provides a conceptual framework for understanding experimental design and it builds a strong foundation for readers who wish to pursue more advanced topics.

The focus throughout the book is conceptual, with our greatest emphasis being on promoting an understanding of the logical underpinnings of design and analysis. This is perhaps most evident in the first part of the book dealing with the logic of design and analysis, which touches on relevant issues in philosophy of science and past and current controversies in statistical reasoning. But the conceptual emphasis continues throughout the book, where our primary concern is with developing an understanding of the logic of statistical methods. This is why we present definitional instead of computational formulas, relying on statistical packages to perform actual computations on a computer. This emphasis allows us to concentrate on the meaning of what is being computed instead of worrying primarily about how to perform the calculation. Nevertheless, we recognize the importance of doing hand calculations on . . .

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.