RE: group-based judgement

From: Smigrodzki, Rafal (SmigrodzkiR@msx.upmc.edu)
Date: Sat May 25 2002 - 12:07:09 MDT


                Charles Hixson [mailto:charleshixsn@earthlink.net] wrote:

                So either everyone gets searched, or the selection of
targets is made by
                some non-involved process (say, throwing a die). Or the
process becomes
                non-random. And you've already said what that causes.

                ### I wonder if you could maximize the effect of your
searches by using a weighting procedure. At first you perform random
searches, to gather statistical data. You might find features associated
with increased or decreased likelihood of carrying the searched-for items.
Then you incorporate a positive or negative weight to each decision to
search any individual (first you appraise the relevant features, compute the
weight, and then throw dice to decide if the person will be searched). This
method would be applied uniformly to all, without discretion given to the
security officer.

                Rafal

                



This archive was generated by hypermail 2.1.5 : Sat Nov 02 2002 - 09:14:21 MST