Re: poly: Are people roughly Bayesian?

Robin Hanson, <hanson@econ.berkeley.edu>, writes:
> >"Broadening the Tests of Learning Models"
> > Using 123 subjects each in 480 trials, we compare five
> > existing learning models plus several variants, including the
> > well-known Bayesian, Fuzzy Logic, connectionist, exemplar and
> > ALCOVE models. We find that subjects do learn to distinguish the
> > symptom configurations, that subjects are quite heterogeneous in
> > their response to the task, and that only a small part of the
> > variation across subjects arises from the differences in
> > treatments.

Saying that the subjects are "quite heterogenous" suggests to me that
some subjects did much better than others.

> > The most striking finding is that the model that
> > best predicts subjects' behavior is a simple Bayesian model with
> > a singe fitted parameter for prior precision to capture
> > individual differences. We use recent rolling regression
> > techniques to elucidate the behavior of this model over time and
> > find some evidence of overresponse to current stimuli.

Would this imply that the "smarter" subjects, the ones who did better
than the others, differed in a specific way, relating to "prior precision"
in a Bayesian learning model? If so this would seem to shed light on
the nature of intelligence. (Or maybe this is just another way of saying
that some people couldn't seem to learn from their mistakes.)

I am not familiar enough with Bayesian learning to know what "prior
precision" means in this context. Can someone explain?

Thanks,
Hal
Received on Fri Jan 8 14:28:03 1999