At trial, witnesses likely have rehearsed their answers to trial questions and believe that they made an accurate identification. These contaminated conditions might undermine the cues observers (jurors) use to judge accuracy. Yet, during the initial lineup, several behavioral cues should facilitate discriminability (e.g., witness confidence/fluency). Over two experiments, we evaluated observer discriminability with a two-phase paradigm. In phase 1, witness-participants (N=300) completed a video-recorded target-present or target-absent lineup, engaged in pretrial rehearsal, received confirmatory post-identification feedback (Experiment 2), and provided video-recorded testimony. In phase 2, we randomly assigned participants (N=700) to a 2(Accuracy: accurate, inaccurate) x 3(Evidence: lineup-video-only, testimony-video-only, lineup-video-plus-testimony-video) mixed-design. Observers judged witness accuracy and rated their confidence in their decision. Though data collection for Experiment 2 is ongoing, signal-detection analyses revealed that discriminability did not differ across evidence conditions and was generally poor. Post-hoc analyses suggest that observers are influenced by reliable (confidence/fluency) and unreliable cues (descriptions).

This study examined how subjective evaluation of memory functioning, verbal description, and type of lineup influence recognition accuracy and confidence in identifying a perpetrator. After watching a robbery video, 228 participants performed a filler task, then either described the robber or completed a control task, followed by a lineup identification (optional choice: target-present or target-absent, or forced choice target absent). Participants also completed metamemory assessments (Eyewitness Metamemory Scale, Memory Distrust Scale, and Squire Subjective Memory Questionnaire). No verbal overshadowing effect was found in any lineup condition. Interestingly, in the optional choice target-absent lineup, those who described the perpetrator were more accurate in indicating the absence of the perpetrator. A positive correlation between confidence and response accuracy was confirmed only for target present force choice lineup. No link was found between subjective memory evaluations and recognition accuracy. However, participants who rated their memory higher were more confident in their answer.

Research suggests that people regularly engage in episodic future thinking (EFT); however little research has examined the impact of EFT on eyewitness recollection. EFT for eyewitness police interviews offer an interesting situation in which people generally have little episodic experiences to draw upon but have a vast amount of semantic content. Therefore, the present study examined the content of simulations of a police interview, and how EFT impacted memory for the crime. After watching a crime video, participants either engaged in EFT, remembered the crime, or completed a filler task, before a final memory test. We found that the experience of imagining a police interview was comparable to imagining a commonly experienced event. Additionally, the amount of crime details retrieved during EFT positively predicted memory accuracy. Our findings suggest that eyewitnesses with no eyewitness experience can simulate robust police interviews that may impact various facets of their memory and experience.

Invoking a stereotype by setting a context at retrieval may distort memory reports to more closely match the stereotype. And such distortions may be more pronounced for ambiguous experiences. In this preregistered study, participants in the role of friends experienced two virtual reality scenarios involving interactions between a heterosexual couple on two consecutive days. These scenarios involved interpersonal conflict (e.g., a discussion about budget) that was either resolved (neutral/positive condition) or unresolved (negative condition developed to convey early stages of domestic abuse). On the third day, participants were interviewed about these scenarios in a role of witnesses. Participants were either informed that a crime (neutral context) or a case of domestic abuse (abuse context) was reported that involved the couple. Preliminary results indicated that under the abusive context instructions, participants reported more abusive content even when the witnessed scenarios were neutral/positive and did not involve any indication of abuse.