Linking Eyewitness Recognition Memory to Identification Accuracy using Item Response Theory

<p>Objective(s): Eyewitness testimony plays an important role in legal proceedings, particularly when other evidence is limited. Although extensive research has examined factors that influence eyewitness memory and eyewitness identification, the relationship between these two aspects of eyewitness evidence remains poorly understood. In the present study, we examined the relationship between eyewitness memory and identification accuracy using an item response theory (IRT) framework.Hypotheses: We predicted that participants who accurately identified the perpetrator would demonstrate higher recognition memory performance on both memory for persons and memory for the witnessed event than participants who identified a filler or made no identification.Method: In a pilot study (N = 175) and a preregistered main study (N = 794), participants watched a mock crime video including two perpetrators from the Jena Eyewitness Research Stimuli (Kruse &amp;amp; Schweinberger, 2023). Participants then completed a 40-item two-alternative forced-choice recognition memory task and made identification decisions from two simultaneous eight-person lineups.Results: The IRT model identified two latent dimensions of eyewitness memory: Person memory and Event memory. Interval-based hypothesis testing indicated a meaningful association between eyewitness identification accuracy and Person memory: The estimated difference in recognition performance exceeded our smallest effect size of interest (1 detail). In contrast, identification accuracy was not meaningfully associated with Event memory.Conclusions: Eyewitness memory and identification are related, but their association depends on the type of memory assessed. Person memory (e.g., perpetrator, victim), but not Event memory (e.g., a theft), was meaningfully associated with identification accuracy. These findings highlight the importance of distinguishing different dimensions of eyewitness memory and considering the psychometric properties of memory measures when examining factors that may undermine eyewitness memory. In particular, researchers should carefully consider which aspect of eyewitness memory their measures assess when drawing conclusions about factors that may affect eyewitness identification accuracy.</p>