Lee Jussim & Jon A. Krosnick 
Research Integrity [PDF ebook] 
Best Practices for the Social and Behavioral Sciences

The primary goal of science is to "get it right, " meaning that scientists seek to accurately document the world as it is. While erroneous conclusions and flawed theories can and do occur, they can only be tolerated as long as reliable mechanisms of self-correction exist, but an array of recent evidence suggests that this is not always the case. This book offers a behavioral science perspective on how scientific practice becomes compromised and provides recommendations for improvement. Broadening the discussion of research integrity beyond replication, publication biases, statistics, and methods, this book addresses the full complexity of the issue and serves academics and policy makers concerned with the reliability and validity of scientific findings across the social sciences. It tackles challenges presented by published reports and textbooks, addresses the ways that institutional review boards (IRBs) can influence the course of research, and discusses the weaknesses of meta-analysis, which is often recommended as a possible corrective measure for suboptimal scientific practice. The book concludes with an organizing framework to investigate how scientists’ behaviors can impact the reliability and validity of scientific research.