The Subtle Science of Misleading with Statistics
summary
Explains concrete ways numbers mislead: swapping arithmetic/geometric/harmonic means to change a result, hiding variance so a 105 ms mean with standard deviation 4.1 vs 45 ms looks equivalent, and cherry‑picking ranges to claim no warming from 1997–2022. Shows a headline 82% claim that really comes from a tiny subgroup in a 22,500 sample, and how a log scale makes 38.3→40.1 look huge. Recommends checking ranges, sample framing, confounders and using an LLM to flag biased questions.
AI-written summary · about
chapters (10)
- 0:00Introduction to statistics
- 3:28The meaning of averages
- 7:26Importance of deviation
- 9:36Sampling and extrapolation
- 12:23Identifying common biases
- 18:49Visual manipulation
- 21:31Misleading data presentation
- 25:35Correlation vs. causation
- 31:30Data dredging and p-hacking
- 36:53Critical thinking and conclusion
related talks
32mThe Subtle Art of Lying with Statistics
27mThe Subtle Art of Lying with Statistics
40mHow to Lie With Stats and Charts
32mIntroducing Reality into Data Metrics
05mPerception and bias and metrics, oh my!
38mKnow your data: The stats behind your alerts
37mHow to Start Critical Thinking in Data Science
32mLearning Statistics Will Save Your Life