Publication bias

What is publication bias?

Researchers are more likely to publish a paper when that paper yields a relatively large treatment effect, and less likely to do so when the study yields a relatively small treatment effect.  This can lead to a situation where the body of published studies is biased, in that it tends to exaggerate the true magnitude of the treatment effect.  These issues are discussed in the book "Publication Bias in Meta-Analysis."

 

Publication Bias in Meta-Analysis: Prevention, Assessment and Adjustments
Hannah Rothstein, Alex J Sutton, Michael Borenstein (eds).
Hardcover, 392 Pages, Wiley, 2005

This book traces the history of the publication bias, explains how to assess the potential impact of bias in any given meta-analysis, and discusses mechanisms that are being adopted in hopes of minimizing the potential for bias in the future. 

This book can be ordered online.

 

Download book chapters

Preface
Table of Contents
Software for meta-analysis and publication bias

Download datasets

Dataset 1 – Teacher expectancy and student intelligence
Dataset 2 – Passive smoking and lung cancer
Dataset 3 – Employment interview performance and job performance

Step-by-step instructions for using various computer programs to address publication bias

Errata and updates

 

Why use CMA software?
image description

Comprehensive Meta-Analysis

Comprehensive Meta-Analysis is a powerful computer program for meta-analysis. The program combines ease of use with a wide array of computational options and sophisticated graphics.