Got perp? Eyewitness accuracy, decision processes, and presentation procedures using sequential lineups

It was my objective to understand whether accurate and inaccurate eyewitnesses could be distinguished by their decision-making during a sequential-lineup. All eyewitnesses, except in Study 6, were shown a videotaped crime and presented with sequential lineups.; Study 1 was designed to identify the decision processes of eyewitnesses. While viewing a culprit-present lineup, witnesses were asked to "think aloud" and later describe in writing their thoughts as they reached a decision for each photograph; five decision process statements were then created or selected from previous research. In Study 2, the main dependent measure asked eyewitnesses to endorse all applicable decision process statements from Study 1. Factor analysis revealed a simple matching strategy containing three decision processes and a deliberative strategy with four decision processes. Accurate eyewitnesses were significantly associated with the simple matching strategy, and inaccurate eyewitnesses with the deliberative strategy.; An automatic recognition statement was added to the decision process statements. Study 3 looked at inaccurate identifications in culprit-absent lineups and found that the decision processes of inaccurate eyewitnesses did not differ regardless of having selected an innocent suspect replacement or a known innocent picture. Study 4a and 4b successfully replicated previous findings using a new set of experimental materials with different witness viewing conditions.; Study 5 demonstrated that accuracy rates could not be predictably influenced via the manipulation of witness decision processes. Witnesses forced to use deliberative decision processes were not subsequently less accurate. Witnesses forced to use simple matching and automatic processes were also not subsequently more accurate. Study 6 participants were asked to postdict witness accuracy. They were given previous eyewitness identification judgment forms and some were informed about the decision strategies found to be indicative of accuracy and some were not. Unexpectedly, informed participants did not outperform the uninformed or perform better than chance.; Studies 7 and 8 tested whether logical modifications to the sequential procedure would affect accuracy. In Study 7, only culprit-present lineups were conducted and seeing it twice before making any identification (no-ID-first-view) presentation produced significantly greater accuracy than the traditional presentation. Study 8 served as a replication and extension, using both culprit-present and culprit-absent lineups. The superiority of the no-ID-first view condition did not reach significance. The implications of Studies 1--8 for memory, face recognition, and the legal system are discussed.