The advent of statistical software for the personal computer in the 1980s can be argued to be both a great advancement as well as a critical liability for the field of research. Statistical software has certainly made difficult analytical tasks easier to accomplish, enabling more people to benefit from the use of quantitative techniques. The ever-increasing speed of the personal computer allows researchers to conduct complex analysis in minutes that would have taken days to complete manually. However, with this increased speed and usability comes pitfalls. The ease and speed of statistical software has encouraged some researchers to take a shotgun approach to analysis by running large numbers of analyses instead of strategically selecting analyses guided by theory. Statistical packages also make it possible to run complex procedures that may be misapplied or misinterpreted by researchers without a solid understanding of statistical principles.
No book can stop unscrupulous researchers from supporting their hypotheses by cherry picking favorable results from hundreds of analyses conducted. Well-intentioned students and researchers, however, can turn to Andy Field's book Discovering Statistics using SPSS for Windows for the statistical background and guidance needed to appropriately select, execute, and interpret results.
Field's book bridges the gap between introductory/intermediate statistical textbook and software manual. This engaging, easy to read book leads the reader through:
· Introduction to statistical models
· Exploring data
· Logistic regression
· Comparing means
· Complex ANOVA
· Repeated measures design
· Exploratory factor analysis.
Each topic of the book begins with an overview of the applicable statistical theory. The theory is presented in non-technical language and references numerous sources for readers wanting a more in-depth review. When background material does become technical, these sections are specially labeled to alert non-technical readers. Field focuses particularly on the statistical assumptions of each statistical technique and how to use SPSS to test for them.
Following the review of statistical theory, each topic includes one or more exercises. Readers are guided step by step through each analysis from dummy coding data, through entering in the necessary SPSS commands, to interpreting SPSS output. An accompanying CD furnishes the data sets used in each exercise, allowing readers to work though each exercise and check results against the book. Through these exercises, Field shows how results can be misinterpreted without a thorough investigation of the data. Great pains are taken throughout the exercises to demonstrate the perils of blindly trusting SPSS output without understanding the theory and underlying statistical assumptions.
From this review, it is obvious that I greatly enjoyed reading the book. I assume that this book is intended primarily for students and practitioners without an extensive background in statistics. However, I found this book to be beneficial in reviewing concepts studied in graduate school more than ten years ago. Field accomplishes the difficult task of simplifying complex topics into everyday language without talking down to his readers. This book is perfectly suited for the non-statistical expert looking for guidance with running analyses with SPSS.