Monday, January 27, 2014

Lars P. Syll — The dilemma of probability theory (wonkish)


This importantly also means that if you cannot show that data satisfies all the conditions of the probabilistic nomological machine, then the statistical inferences used – and a fortiori neoclassical economics – lack sound foundations!
The dilemma of probability theory (wonkish)
Lars P. Syll | Professor, Malmo University

4 comments:

Anonymous said...

It seems to me that Lars is assuming a frequentist reading of probability statements within economics, but that his critique does not apply to a Bayesian reading.

Tom Hickey said...

Lars has also posted contra Bayesianism, too. Search on Syll" and "Bayesianism."

See for instance, Objections to Bayesian statistics. Be sure to read the comments too.

Clint Ballinger said...

This may be of interest to some "Why inferential statistics are inappropriate for development studies and how the same data can be better used"
http://mpra.ub.uni-muenchen.de/29780/

Clint Ballinger said...

Dan - these critiques apply to inferential stats in general, both frequentist and Bayesian.
As I have discussed elsewhere, what is really needed is much better use of descriptive statistics.