Multiple Regression with Discrete Dependent Variables

Paperback | March 25, 2009

byJohn G. Orme, Terri Combs-orme

not yet rated|write a review
Most social work researchers are familiar with linear regression techniques, which are fairly straightforward to conduct, interpret, and present. However, linear regression is not appropriate for discrete dependent variables, and social work research frequently employs these variables,focusing on outcomes such as placement in foster care or not; level of severity of elder abuse or depression symptoms; or number of reoffenses by juvenile delinquents in the year following adjudication. This book presents detailed discussions of regression models that are appropriate for a variety of discrete dependent variables. The major challenges of such analyses lie in the non-linear relationships between independent and dependent variables, and particularly in interpreting and presentingfindings. Clear language guides the reader briefly through each step of the analysis, using SPSS and result presentation to enhance understanding of the important link function. The book begins with a brief review of linear regression; next, the authors cover basic binary logistic regression, whichprovides a foundation for the other techniques. In particular, comprehension of the link function is vital in order to later interpret these methods' results. Though the book assumes a basic understanding of linear regression, reviews and definitions throughout provide useful reminders of importantterms and their meaning, and throughout the book the authors provide detailed examples based on their own data, which readers may work through by accessing the data and output on companion website. Social work and other social sciences faculty, students, and researchers who already have a basicunderstanding of linear regression but are not as familiar with the regression analysis of discrete dependent variables will find this straightforward pocket guide to be a terrific boon to their bookshelves.

Pricing and Purchase Info

$29.95

Ships within 1-3 weeks
Ships free on orders over $25

From the Publisher

Most social work researchers are familiar with linear regression techniques, which are fairly straightforward to conduct, interpret, and present. However, linear regression is not appropriate for discrete dependent variables, and social work research frequently employs these variables,focusing on outcomes such as placement in foster ca...

John Orme and Terri Combs-Orme are Professors of Social Work at the University of Tennessee.

other books by John G. Orme

Format:PaperbackDimensions:256 pages, 8.25 × 5.5 × 0.68 inPublished:March 25, 2009Publisher:Oxford University PressLanguage:English

The following ISBNs are associated with this title:

ISBN - 10:0195329457

ISBN - 13:9780195329452

Look for similar items by category:

Customer Reviews of Multiple Regression with Discrete Dependent Variables

Reviews

Extra Content

Table of Contents

Preface1. Introduction to Regression Modeling2. Regression with a Dichotomous Dependent Variable3. Regression with a Polytomous Dependent Variable4. Regression with an Ordinal Dependent Variable5. Regression with a Count Dependent VariableAppendix A: Description of Data SetsAppendix B: LogarithmsGlossary