The erroneous signals of detection theory

Pete C Trimmer, Sean M Ehlman, John M McNamara, Andrew Sih

Research output: Contribution to journalArticle (Academic Journal)peer-review

17 Citations (Scopus)

Abstract

Signal detection theory has influenced the behavioural sciences for over 50 years. The theory provides a simple equation that indicates numerous 'intuitive' results; e.g. prey should be more prone to take evasive action (in response to an ambiguous cue) if predators are more common. Here, we use analytical and computational models to show that, in numerous biological scenarios, the standard results of signal detection theory do not apply; more predators can result in prey being less responsive to such cues. The standard results need not apply when the probability of danger pertains not just to the present, but also to future decisions. We identify how responses to risk should depend on background mortality and autocorrelation, and that predictions in relation to animal welfare can also be reversed from the standard theory.

Original languageEnglish
JournalProceedings of the Royal Society B: Biological Sciences
Volume284
Issue number1865
DOIs
Publication statusPublished - 18 Oct 2017

Keywords

  • Journal Article

Fingerprint Dive into the research topics of 'The erroneous signals of detection theory'. Together they form a unique fingerprint.

Cite this