Fair Profiling
Abstract
There are several strategies available to police “stopping” suspects. Most efficient
is to stop only members of the group with the highest a priori probability of guilt;
least efficient is indiscriminate stopping. An efficient option that satisfies one
criterion for fairness is a strategy that matches stop probability to risk probability.
But a strategy that chooses stop probabilities so that the absolute number of innocents
stopped is equal for all groups is close to maximally efficient and seems fair by
almost any criterion.
Type
Journal articlePermalink
https://hdl.handle.net/10161/2883Collections
More Info
Show full item recordScholars@Duke
John E. R. Staddon
James B. Duke Distinguished Professor Emeritus of Psychology and Neuroscience
Until my retirement in 2007, my laboratory did experimental research on learning and
adaptive behavior, mostly with animals: pigeons, rats, fish, parakeets. We were particularly
interested in timing and memory, feeding regulation, habituation and the ways in which
pigeons and rats adapt to reward schedules. The aim is to arrive at simple models
for learning that can help to identify the underlying neural mechanisms. I continue
to do theoretical and historical work on the power law in
Material is made available in this collection at the direction of authors according to their understanding of their rights in that material. You may download and use these materials in any manner not prohibited by copyright or other applicable law.
Rights for Collection: Research and Writings