Clearer Thinking with Spencer Greenberg Statistics Intuitions and Social Science Reproducibility (with Stuart Buck)
Jul 7, 2021
Stuart Buck, former Vice President of Arnold Ventures, discusses the urgent need for reform in math education, advocating for a focus on statistics and practical skills. He tackles the confusion surrounding p-values, emphasizing their misinterpretations in public discourse. Buck explores the repercussions of the reproducibility crisis in social sciences, highlighting the significance of publishing null results. He stresses the importance of rigorous research practices to ensure reliability and transparency in scientific findings. Tune in for insights that could change how we approach education and research!
AI Snips
Chapters
Transcript
Episode notes
Statistical Significance
- The phrase "statistically significant" is misleading and creates a false dichotomy.
- Evidence is a continuum, and p-values close to 0.05 should not be interpreted as absolute proof or disproof.
Publishing Null Results
- Publish null results for theoretically interesting questions or those challenging existing beliefs.
- Avoid publishing null results for random ideas with low prior probability.
Reproducibility and Philanthropy
- The Arnold Foundation became interested in reproducibility due to concerns about evidence-based policy.
- Unreliable research and publication bias can mislead philanthropic decisions.