This preview shows page 1. Sign up to view the full content.
Unformatted text preview: ” “Sick” State of the
World Healthy 2 1 Sick 1 2 In this case, the value of the two correct decisions
must be the same, and the value of the two
incorrect decisions must be the same.
(That way they cancel out.)
(That β = p(H)/p(S)*[v(CR) + v(FA)]/[v(H) + v(M)]
= .8/.2*(2 + -1)/(2 + -1)
=4 0.45 “Healthy” 0.4 “Sick” 0.3
0.25 Sick Healthy 0.2
0.05 # Goodies in the Blood 14
0 85 70 0
55 Probability of Disease 0.35 The optimal location for the response
threshold depends on the objective of
the decision maker. What goal does he
or she want to accomplish (maximize
expected value of the decision,
maximize the number of correct
decisions, SDT has widespread applications:
deciding who would default on a loan
deciding whether to hire a worker
determining whether someone is lying
and any other decision that is made
repeatedly based on ambiguous
evidence Another Example: Jury Decision Making Probability “Acquit” “Convict”
Guilty Innocent P erceived Culpability What are acceptable levels of errors?
Convicting the innocent - False Alarms
Acquitting the guilty - Misses
Within this framework, how good must
jurors be to achieve reasonable rates of
error Assuming normal dis...
View Full Document
This document was uploaded on 02/15/2014 for the course PSYCH 002 at Community College of Philadelphia.
- Fall '14