Eyewitness identification: Bayesian information gain, base-rate effect equivalency curves, and reasonable suspicion

Area of Science:

• Cognitive Psychology
• Forensic Science
• Legal Psychology

Background:

• Eyewitness identification is a critical but flawed component of the justice system.
• Traditional research focuses on system variables like instructions and filler similarity.
• The influence of base rates on identification accuracy is often underestimated.

Purpose of the Study:

• To apply a novel Bayesian framework to analyze eyewitness identification.
• To quantify the impact of system variables and base rates on identification reliability.
• To explore the implications for legal standards in identification procedures.

Main Methods:

• Bayesian analysis incorporating base rates as a system variable.
• Utilized prior-by-posterior and information-gain curves on published experimental data.
• Introduced Base Rate Effect-Equivalency (BREE) curves for comparative analysis.

Main Results:

• Information-gain curves demonstrate that lineups are more effective at implicating the guilty than exonerating the innocent.
• System variables and witness confidence moderate information-gain curves.
• BREE curves show that base rate changes have a greater impact than traditional system variables.

Conclusions:

• The base rate is a crucial, controllable system variable in eyewitness identification.
• Modest adjustments to base rates can significantly enhance identification reliability.
• Findings suggest a re-evaluation of reasonable-suspicion criteria for identification procedures.