U.S. flag

An official website of the United States government, Department of Justice.

NCJRS Virtual Library

The Virtual Library houses over 235,000 criminal justice resources, including all known OJP works.
Click here to search the NCJRS Virtual Library

Application of Signal Detection Theory to Decision-Making in Forensic Science

NCJ Number
187611
Journal
Journal of Forensic Sciences Volume: 46 Issue: 2 Dated: 2001 Pages: 294-308
Author(s)
Victoria L. Phillips Ph.D.; Michael J. Saks Ph.D.; Joseph L. Peterson DCrim
Date Published
March 2001
Length
15 pages
Annotation
Signal Detection Theory (SDT), a product of the marriage of mathematical statistics and advances in electronic communications in the 1950's, constitutes both a body of knowledge and a set of analytical methods designed to rigorously examine decision making by machines and people alike; this paper is intended to make SDT more accessible to forensic science.
Abstract
Although SDT is widely used in research in fields as varied as aviation, psychiatry, and radiology, to date it has barely been used in forensic science. After reviewing the basic concepts of SDT, this paper discusses various aspects of forensic science theory and practice that likely could benefit from the application of SDT; SDT analysis is illustrated by using forensic science proficiency data. SDT is particularly well-suited to the study of decision making where the presence or absence of a "signal" has to be discerned amid a complex array of other, ambiguous stimuli. SDT would facilitate numerous informative lines of research in forensic science; for example, from existing forensic science proficiency data, clearer answers could be found concerning the performance of participants in those studies. Most notably, raw diagnostic skill could be distinguished from the confounding effects of decision thresholds. The relative accuracy produced by various protocols and technologies could be more clearly evaluated. The abilities of examiners with different backgrounds, training, and experience could be more completely and clearly assessed. The forces that cause decision thresholds to rise, to fall, or to remain stable could be identified, and optimal decision thresholds could be more rationally and effectively determined. 4 figures, 1 table, and 74 references

Downloads

No download available

Availability