A Bayesian decision-theoretic approach appears to me as a sensible idealization of a guide to behaviour. At the same time i would like to understand why my behaviour is not always of this form: I sometimes use randomization and I sometimes find confidence intervals acceptable. Not all of my problems have an explicit cost function. Am I lazy or irrational? Do I use non-Bayesian conventions to help communicate? Is the cost of rationality-computation missing from the Bayesian model?
@article{urn:eudml:doc:40830,
title = {On not being rational.},
journal = {Trabajos de Estad\'\i stica e Investigaci\'on Operativa},
volume = {31},
year = {1980},
pages = {321-328},
language = {en},
url = {http://dml.mathdoc.fr/item/urn:eudml:doc:40830}
}
Savage, I. Richard. On not being rational.. Trabajos de Estadística e Investigación Operativa, Tome 31 (1980) pp. 321-328. http://gdmltest.u-ga.fr/item/urn:eudml:doc:40830/