First the link to this week’s complete list as HTML and as PDF.
***
As Damian & Roberts point out a result can be both statistically significant and practically meaningless – something obviously not taught in any sociology department in the world.
***
In their figure 4 Roffet-Salque et al. supply the perfect example of how not to construct isolines. The area of modern France is covered by no less than eight zones with complicated shapes, equalling at least 50 spurious mathematically generated values of pseudodata, all generated out of five or less measured data points. Another criterion is, that maps of isolines must not change their basic form and shape by taking out single data points. The wholly spurious sharply rectangular area in Southern Germany is an example of the opposite. Maps like that are meaningless.
***
Reading Scharl’s review of O’Brien reminded me of something. The deforestation around copper mines has consistently been shown to be less than expected. Working backwards from the rate of regrowth and the amount of charcoal available, we can arrive at an upper limit for the amount of metal extracted and it will probably turn out lower than previous estimates.