The Essence of Multivariate Thinking: Basic Themes and Methods (Multivariate Applications Series) - Hardcover

Harlow, Lisa L.

Hardcover

ISBN 10: 0805837299 ISBN 13: 9780805837292

Publisher: Psychology Press, 2005

Synopsis

The Essence of Multivariate Thinking is intended to make multivariate statistics more accessible to a wide audience. To encourage a more thorough understanding of multivariate methods, author Lisa Harlow suggests basic themes that run through most statistical methodology. The most pervasive theme is multiplicity. The author argues that the use of multivariate methods encourages multiple ways of investigating phenomena. She explains that widening our lens to identify multiple theories, constructs, measures, samples, methods, and time points provide greater reliability and validity in our research. Dr. Harlow then shows how these themes are applied to several multivariate methods, with the hope that this will ease understanding in the basic concepts of multivariate thinking. Formulas are kept at a minimum.

The first three chapters review the core themes that run through multivariate methods. Seven different multivariate methods are then described using 10 questions that illuminate the main features, uses, multiplicity, themes, interpretations, and applications. The seven methods covered are multiple regression, analysis of covariance, multivariate analysis of variance, discriminant function analysis, logistic regression, canonical correlation, and principal components/factor analysis. The final chapter pulls together the principal themes and features charts that list common themes and how they pertain to each of the methods discussed.

The Essence of Multivariate Thinking, features:

• A unique focus on the underlying themes that run through most multivariate methods.
• A dual focus on significance tests and effect sizes to encourage readers to adopt a thorough approach to assessing the significance and magnitude of their findings.
• A detailed example for each method to delineate how the multivariate themes apply.
• Tabular results from statistical analysis programs that mirror sections of the output files.
• A common dataset throughout the chapters to provide continuity with the variables and research questions.
• A CD with data, SAS program setup and output, homework exercises, and chapter lectures.

This book is useful to advanced students, professionals, and researchers interested in applying multivariate methods in such fields as behavioral medicine, social, health, personality, developmental, cognitive, and industrial-organizational psychology, as well as in education and evaluation. A preliminary knowledge of basic statistics, research methods, basic algebra, and finite mathematics is recommended.

About this title

Review

The most unique aspect of Essence is a focus on a series of themes that are later applied to individual statistical procedures. This approach is very effective as it gives readers a sense that they are learning about how different applications fit a general model of multivariate statistics rather than focusing on applications in isolation....it is a very good text for teaching....how easy it would be to build a course around the book and supporting materials. The text's strengths are its clear and explicit focus on thinking about topics and its highly readable presentation.
—PsycCRITIQUES

In view of the ever expanding applications of multivariate statistics in various disciplines...books that make the topic of multivariate statistics more accessible and comprehensible to a wide audience are welcome additions....This book fits this need well and does not require knowledge of advanced mathematical methods beyond basic algebra and finite mathematics. Thebook presumes knowledge of basic statistics and methods as taught at the undergraduate level in most social science fields....could be covered in a statistics course for first-year graduate students or advanced undergraduate. One noteworthy feature of this book is its organization. The author suggests basic themes that run through most statistical methodology and then show how these themes are applied to several multivariate methods. 
—Technometrics

Edition details

• Publisher: Psychology Press
• Publication date: 2005
• ISBN 10: 0805837299
• ISBN 13: 9780805837292
• Binding: Hardcover
• Language: English
• Edition number: 1
• Number of pages: 264