Repligate: Reliability and Reproducibility in Psychology Syllabus

Abstract

This seminar will address issues relevant to the current controversy over the reliability of psychological research, a controversy which seems to be white-hot right now. Topics will include (but not be limited to)

• Critiques of the current science, including the claim that “most published research findings are false”
• Controversies over the replicability of particular findings, including behavioral priming and ESP
• The way practices by journal editors, granting agencies, and hiring committees do and do not encourage reliable, replicable research
• Recommendations for improving conduct and reporting of research, including statements by professional societies, journals, and government agencies
• Related methodological issues including o Null-hypothesis statistical testing oThe “new statistics” emphasizing effect sizes and confidence intervals o Exploratory vs. confirmatory research o p-curving, the test for excess significance, and other statistical tools intended to detect questionable research
• Defenses of the current state of psychological research, and the various kinds of push back against what some call the “anti-false-positives movement” and others simply call “shameless little bullies.” Readings will include journal articles, editorials, blog posts and probably even a tweet or two. While a lengthy reading list is provided, we will be selective in what we actually read in depth; skimming will be encouraged when appropriate. The course structure will be discussion of the readings. That’s all. No exams or papers. But do come to class ready to talk.