fullROC: An R package for generating and analyzing eyewitness-lineup ROC curves.

Area of Science:

• Psychology
• Forensic Science
• Criminal Justice

Background:

• Police lineups are crucial for eyewitness identification but traditionally focus on suspect identification, neglecting exculpatory evidence.
• Past research has primarily analyzed inculpatory behaviors, overlooking the value of filler identifications and rejections in determining innocence.

Purpose of the Study:

• To introduce the fullROC R package for analyzing police lineup data using a comprehensive Receiver Operator Characteristic (ROC) curve method.
• To enable researchers to incorporate all lineup outcomes (suspect identifications, filler identifications, and rejections) into a single analytical framework.

Main Methods:

• Development of the fullROC package for the Comprehensive R Archive Network (CRAN).
• Utilizing Receiver Operator Characteristic (ROC) curves to analyze full lineup data, encompassing both inculpatory and exculpatory evidence.
• Functions for adjusting identification rates, generating ROC curves, calculating Area Under the Curve (AUC), and comparing AUCs of different lineups.

Main Results:

• The fullROC package provides tools to analyze lineup data by incorporating all possible outcomes.
• The Area Under the full ROC curve (AUC) quantifies the lineup procedure's capacity to discriminate between guilty and innocent suspects.
• Demonstrated functionality using both simulated and empirical eyewitness identification data.

Conclusions:

• The fullROC package offers a valuable tool for eyewitness researchers to enhance the analysis of lineup data.
• This approach provides a more complete understanding of eyewitness performance by considering both evidence of guilt and innocence.
• Facilitates statistically robust comparisons of different lineup procedures.