Monday, August 18, 2014

My Year of Darwin: 8/18/2014 Darwin does stats

 Charles Darwin

"Now Fries has remarked in regard to plants, and Westwood in regard to insects, that in large genera the amount of differences between the species is often exceedingly small. I have endeavoured to test this numerically by averages, and, as far as my imperfect results go, they confirm the view" Darwin, Origin of Species 2nd edition

I added this passage because this is the first time in the Origin that Darwin does a statistical test - although nothing like a statistical test we use today. Unfortunately, my version does not include a table but I suspect he groups insects as either in large genera with many species and small genera with few species. The lack of a statistical test with p-values is no fault of Darwin - they did not exist until the 1930's. 

And p-values should die (and substituted with Bayesian posterior probs). 

No comments:

Post a Comment