From: Damien Broderick (email@example.com)
Date: Sun Aug 27 2006 - 14:25:39 MDT
At 03:08 PM 8/27/2006 -0400, Jeff Medina wrote:
>the example that started this
>subthread of the argument (the feminist bank teller question and the
>typical responses it gets) *is a canonical heuristics-and-biases
Everyone knows that.
>That's where this aspect of the debate began; Eliezer is
>not trying to steer away from modeling, except insofar as Richard
>keeps seeming to think modeling is part of the heuristics-and-biases
>subfield, which it isn't.
I think Richard was trying for a reframing of the example, as a way
of demonstrating how a quick&dirty fast&frugal catch-as-catch-can
structure like a typical brain/body response actually operates, in
the list's context of seeking insight into machine analogues of
cognition, evaluation and response.
>Richard posed what he thinks is a more plausible alternative
>explanation the feminist bank teller experimental results. That
>alternative was one of the first, most obvious alternatives proffered
>by psych folks, and it has been experimentally tested and *does not*
>explain people's responses to the feminist bank teller problem.
It seems to me quite consonant with the (experimentally demonstrable)
Roschean model of type representation. Eliezer noted "the standard
experimental result *and its standard interpretation*: subjects
substitute judgment of representativeness for judgment of
probability... Judgment by representativeness, like judgment by
availability, is a heuristic that can often produce useful results -
some even say a fast and frugal heuristic - but the heuristic also
produces systematic biases, which are what reveal to us that the
heuristic exists in the first place, and what pin[s] down how the
heuristic must be working to produce those specific biases." Yes,
yes, yes, nor did Richard dispute the fallibility *for some rather
contrived purposes* of templated heuristics that are effective *for
other daily purposes*. So why all the bad temper and foot-stamping?
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:57 MDT