Police routinely give eyewitnesses multiple opportunities to identify the same suspect, and numerous exoneration cases demonstrate that this practice can contribute to wrongful convictions. Empirical research addressing this practice shows it can lead to the repeated-suspect effect, which is a significant increase in suspect identifications after the same suspect has been presented in a previous showup or lineup (Steblay & Dysart, 2016). Procedures that tend to increase the chance of innocent suspect identifications are considered suggestive and produce unreliable eyewitness decisions. Thus, the use of multiple identification opportunities is considered suggestive and are discouraged by researchers (Wells et al., 2020). Despite this, eyewitness testimony obtained using suggestive procedures is frequently used at trial anyway because it is still admissible in court if other criteria are met indicating the identification was “nevertheless reliable” (Manson v. Brathwaite, 1977; Wells & Quinlivan, 2009). This dissertation builds on past research in this area by examining the effect of more the one intervening lineup and biased intervening lineups containing the same innocent suspect in two experiments, and how these different intervening lineup manipulations impact identification outcomes, confidence, and mechanism-related questions.

Participants in both experiments watched a crime video, completed an intervening task phase, then evaluated a final, fair lineup. For Experiment 1, the intervening tasks were a fair lineup, a biased lineup, or a reading comprehension control task, and the final lineup contained either an innocent suspect, repeated from the intervening lineup, or the true culprit embedded among five fillers. For both studies, a repeated-suspect effect occurred such that a fair intervening lineup containing the innocent suspect resulted in more misidentifications of that same innocent suspect in the final lineup, consistent with past research. Furthermore, when participants received an intervening lineup that was biased towards the repeated innocent suspect, the repeated-suspect effect was more pronounced with significantly more innocent suspect picks from the final lineup when compared with participants who received a fair intervening lineup. Similarly, in Experiment 2, presenting participants with two intervening lineups exacerbated repeated-suspect effects relative to only one intervening lineup, particularly when one of those intervening lineups was biased. In Experiment 1, participants who received a final lineup containing the culprit were equally able to identify the culprit regardless of the intervening task condition, indicating that the memory for the true culprit was maintained despite exposure to misleading intervening lineups.

Overall, these studies confirm that multiple identification attempts that each contain the same innocent suspect decreases the reliability of eyewitness decisions and increases the risk of misidentification for the innocent suspect. Moreover, these effects become stronger when the intervening lineups are biased or when more intervening lineups are introduced. Post-identification questions were also used to determine how the repetition manipulation and lineup bias manipulation might influence the relative contribution of two cognitive (dual-process recognition and source misattribution) and two social (commitment and demand characteristics) mechanisms for repeated-suspect effects. These self-report measures were used to speculate about how these processes are involved in creating repeated-suspect effects and to encourage future research addressing non-memory mechanisms in eyewitness identification research.