You are here

Sage online ordering services are currently unavailable due to network disruption. We are working to restore service as soon as possible. In the meantime, our books are still available widely via retailers and online. Inspection copies can be requested direct from Vitalsource.
If you need further assistance please visit our Contact us page for further information. Thank you for your patience and we apologise for the inconvenience.
Linear Regression

Linear Regression
An Introduction to Statistical Models

  • Peter Martin - University College London, UK, Lecturer in Applied Statistics in the Department of Applied Health Research at University College London.
Additional resources:

May 2022 | 200 pages | SAGE Publications Ltd
Part of The SAGE Quantitative Research Kit, this text helps you make the crucial steps towards mastering multivariate analysis of social science data, introducing the fundamental linear and non-linear regression models used in quantitative research. Peter Martin covers both the theory and application of statistical models, and illustrates them with illuminating graphs, discussing:

·       Linear regression, including dummy variablesand predictor transformations for curvilinear relationships

·       Binary, ordinal and multinomial logistic regression models for categorical data

·       Models for count data, including Poisson, negative binomial, and zero-inflated regression

·       Checking model assumptions and the dangers of overfitting

What is a statistical model
Simple linear regression
Assumptions and transformations
Multiple linear regression: A model for multivariate relationships
Multiple linear regression: Inference, assumptions, and standardization
Where to go from here

Martin provides a comprehensive account of linear regression and offers a detailed and practical guide on how to interpret all the coefficients and statistics included in a model - a valuable resource for social scientists at all stages in their careers.

Jane Elliott
Professor of Sociology at the University of Exeter

The first five chapters set up a clear and solid foundation for understanding statistical models covering a clear explanation of linear regression and its assumptions, the indicators of model fit and predictive power, methods for comparing models with one another as well as complicated cases involving interactions and transformed predictor variables. The final chapter, named ‘Where to Go From Here’, suggests some ways in which the reader could deepen their knowledge of regression, and includes the exploration of some paths that could be taken when/if linear regression is not a suitable model. This book is clearly written and accessible to anyone who has previous basic knowledge of descriptive and inferential statistics. Not only does it include flawless text and graphical explanations, but it is also linked with a support website that supplies data sets for most of the examples used. A big plus is the companion examples/exercises for the open-source software R.

Antonella Cirasola
Clinical Psychologist

This is an excellent introductory text to multivariate analysis of data and is written in accessible language. This text introduces linear regression in a way that is accessible for those with knowledge of descriptive and inferential statistics. The text brings statistical modelling to life while capturing the messiness and ambiguity we may face when interpreting real data. It is engaging and easy to follow. I would highly recommend this for social scientists with an interest in linear regression.

Dr Sally O'Keeffe
City, University of London

This is a must-have resource for people looking for a clear and complete overview of linear regression. There are many books on the topic but Peter Martin’s Linear regression: an introduction to statistical models is among the few that provided me with a crystal-clear explanation of the technique with real research examples. Additionally, the book deals in detail with an often-overlooked aspect of this type of regression: its assumptions. Running a model can be straightforward, but the author is right to remind us that the results can be misleading if the assumptions of the technique are not assessed. The engaging narrative makes this book welcoming to those without a solid statistical background, but it is still able to provide very relevant insights for the more mathematically inclined.

Eliazar Luna
University College London

Sage College Publishing

You can purchase this book and request an instructor sample on our US College site:

Go To College Site

This title is also available on SAGE Research Methods, the ultimate digital methods library. If your library doesn’t have access, ask your librarian to start a trial.