How Correlation of Subsections Will Be Lower Than the Total (In Absolute Value)

 Nassim Nicholas Taleb 

As Paul Krugman would say, the below is pretty wonkish.

In the below clip, Nassim Nicholas Taleb discusses the very important observation that the correlation of subsections will be lower than the total (in absolute value). With implications for "empirical" research.

I'm guessing that many will understand the general gist from the context of the video even if they don't understand all the technicalities.



BTW, the best book I have come across on the abuse of correlations is The Cult of Statistical Significance: How the Standard Error Costs Us Jobs, Justice, and Lives by Stephen T. Ziliak and Deirdre N. McCloskey.

The book is very well written, easy to follow, and provides examples from across the scientific spectrum of abuses, distortions and errors of applied correlations. There are examples in the book from the work of Noble prize winning economist Gary Becker to Merck pharmaceutical claims.



-RW