". . . the writing makes this book interesting to all levels of students. Bobko tackles tough issues in an
easy way but provides references for more complex and complete treatment of the subject. . . . there is a familiarity
and love of the material that radiates through the words."
--Malcolm James Ree, ORGANIZATIONAL RESEARCH METHODS, April 2002
"This book provides one of the clearest treatments of correlations and regression of any statistics book I
have seen. . . . Bobko has achieved his objective of making the topics of correlation and regression accessible
to students. . . . For someone looking for a very clearly written treatment of applied correlation and regression,
this book would be an excellent choice."
--Paul E. Spector, University of South Florida
"As a quantitative methods instructor, I have reviewed and used many statistical textbooks. This textbook
and approach is one of the very best when it comes to user-friendliness, approachability, clarity, and practical
utility."
--Steven G. Rogelberg, Bowling Green State University
Submitted By Publisher, January, 2003
Summary
Correlation and Regression attempts to take statistical theory in correlation and regression and make
it accessible to readers using words, equations, and a variety of applied examples. The examples help explain how
the techniques work and under what circumstances some creativity in application is necessary. The revision keeps
many of the prior classic examples, while adding even more up-to-date examples relevant to the next century of
social science.
The book�s goal is practical; the tone is commonsensical. It is based upon the belief that one can be clear yet
remain precise and accurate. The notions of practicality and common sense reflect the underlying philosophy that
the application of correlation and regression is not necessarily straightforward, and it requires thinking about
each technique in creative ways.
The first half of the book covers topics in correlation; the second half is devoted to regression. Although some
chapters cover traditional topics, other chapters are uniquely conceived as examples of important applications
issues. Chapter IV is devoted to the ubiquitous issues underlying �measurement� in the social sciences. Chapter
V is devoted to the problem of �range restriction.� Chapter VII focuses on specific issues in bivariate regression
(e.g., regression to the mean; utility analysis). Chapters IX and X give more extended, important coverage of interactive
models and validity shrinkage than typically found. In sum, the book is a blend of statistics, design, and measurement
appropriate for applied social science contexts.
Table of Contents
Preface
I. An Introduction, An Overview, And Some Reminders
II. A Review of the Correlation Coefficient and Its Properties
III. Testing Correlations For Statistical Significance
IV. Applications Of Pearson Correlation To Measurement Theory
V. Range Restriction
VI. "Simple," Two-Variable Regression
VII. Three Applications Of Bivariate Regression: Utility Analysis, Regression To The Mean, Partial Correlation
VIII. Multiple (Mostly Trivariate) Regression
IX. Expanding The Regression Repertoire: Polynomial and Interaction Terms
X. More About Regression, And Beyond