Inconclusives and error rates in forensic science: a signal detection theory approach. (28th June 2022)

• Record Type: Journal Article
• Authors: Arkes, Hal R; Koehler, Jonathan J
• Is Part Of: Law, probability & risk. Volume 20:Number 3(2021)
• Page Start: 153
• Page End: 168
• Publication Date: 2022-06-28
• Subjects: signal detection theory -- forensic science -- inconclusive -- error rates; Proximate cause (Law) -- Periodicals; Risk -- Periodicals; Law -- Mathematical models -- Periodicals; Law -- Methodology -- Periodicals; Probabilities -- Periodicals; Risk assessment -- Periodicals; Law -- Mathematical models; Law -- Methodology; Probabilities; Proximate cause (Law); Risk; Risk assessment; Periodicals; 340.1
• DOI: 10.1093/lpr/mgac005
• Languages: English
• ISSNs: 1470-8396

Abstract

Abstract: There are times when a forensic scientist may not be comfortable drawing a firm conclusion about whether a questioned sample that appears to contain useful identifying information did or did not come from a particular known source. In such cases, the forensic scientist may call the sample pair 'inconclusive'. We suggest that signal detection theory (SDT), which is concerned with the detection of weak signals in noisy environments, provides a useful framework for understanding the role that inconclusives play in the various feature-matching forensic sciences. SDT shows that 'inconclusive' is often an appropriate response depending on both the strength of the signal in the samples and the thresholds adopted by the examiner. We also argue that inconclusives should not be coded as either correct or incorrect when tabulating forensic error rates.