Learning Bayesian Statistics How To Get Into Causal Inference
Jan 17, 2024
Dive into the fascinating world of causal inference, where Bayesian concepts merge seamlessly with interactive learning. Discover the significant contributions of Judea Pearl and unravel the complexities of collider bias in epidemiology. Explore the low birth weight paradox, revealing a surprising twist in maternal smoking outcomes. This discussion promises to illuminate how causal diagrams clarify sampling biases and deepen understanding of these intricate statistical relationships.
AI Snips
Chapters
Transcript
Episode notes
Causal Inference Learning
- Build on Bayesian foundations when learning causal inference.
- Add causal inference techniques like regression discontinuity design and difference-in-differences.
Recommended Reading
- Use Judea Pearl's "Book of Why" to learn causal inference.
- The book effectively progresses from Bayesian nets to causal inference concepts.
Low Birth Weight Paradox
- Allen Downey discusses the low birth weight paradox in his new book.
- The paradox confused researchers due to collider bias misinterpreting maternal smoking's impact on low birth weight babies.