Episode 120 - The Reopening: Confounders, Colliders, and Simpson’s Paradox
Max makes his case on the ineffectiveness of lockdowns. Concepts of the week include confounders and colliders, which are subjects in causality theory, and Simpson’s Paradox where correlations flip based on how the data is bucketed.
Sponsor
If you need to export, import, store or exchange large amounts of data on a regular basis, then you can benefit from a compact, fast and versatile data format.
Learn More about RION, Learn about Nanosai
Links
New York City: Covid Data from the City Government
Gothamist: Coronavirus Statistics for New York
New York Post: End New York City’s Lockdown Now!
De Blasio: We’ve seen other places do that and they paid Dearly
New York Times: Coronavirus Deadly in Nursing Homes
New York Times: “Counterfactual Simulation” Says 36K Deaths caused by delay in Lockdown
Columbia University Mailman School of Public Health: Paper on the Counterfactual Simulation
Foursquare: Recovery Index provides Raw Data to Track Cities Coming Back to Life
Tweet from HaraldofW: Are under 65 deaths in Sweden Statistically Insignificant?:
Wikipedia: Simpson’s Paradox
Tweet from boriquagato: No Predictive Power of distancing response on viral spread
Also, State rankings show no pattern
William Briggs: There Is No Evidence Lockdowns Saved Lives.
Norman Fenron, Queen Mary University: Why most studies into COVID19 risk factors may be producing flawed conclusions - and how to fix the problem
- Covers difference Between Confounders and Colliders
More on Colliders, Confounders, and Causality: Judea Pearl’s Book of Why
Time of Israel: Prof Stands by Statement that Pandemic Plays out in 70 Days
