You Are Not So Smart 100 - The Replication Crisis
23 snips
Apr 20, 2017 
Brian Nosek, executive director of the Center for Open Science and a leading voice in the reproducibility movement, dives into the hot-button issue of the replication crisis in psychology. He highlights how much psychological research, particularly around ego depletion, fails replication tests. Nosek discusses publication bias, P-hacking, and the need for reform to bolster trust in science. With a mix of humor and insight, he emphasizes the importance of transparency and the value of questioning our scientific beliefs.
AI Snips
Chapters
Transcript
Episode notes
The File‑Drawer Problem
- Publication bias (the file‑drawer effect) skews the literature toward positive results.
- Unpublished null results make whole fields look stronger than they really are.
The 100‑Study Replication Effort
- A 2015 reproducibility project attempted 100 studies from 2008 and only ~36% replicated.
- That high failure rate pushed the replication issue into public view as a major concern.
How P‑Hacking Produces False Positives
- P‑hacking involves making analytic choices to force significant p‑values.
- These researcher degrees of freedom inflate false positives and mislead journals and readers.