Summary
"Eyewitnesses‘ descriptions of suspects often refer to distinctive facial features, such as tattoos or scars, and the police have to decide how best to create fair lineups in these circumstances. This issue, despite its importance, has attracted insufficient attention in the eyewitness identification literature. Informed by the Police and Criminal Evidence Act code of practice and current police practice, I conducted an empirical evaluation of the different lineup techniques that investigators currently use for suspects with distinctive features.To ensure that a suspect does not stand out because of his distinctive feature,and also to extract more information from the eyewitness, the police either replicate the distinctive feature across all foils in the lineup or conceal the distinctive feature on the face of the suspect. These techniques were tested either in a crossover recognition-memory paradigm (Study 1), or in a lineup-identification paradigm(Studies 2, 3, and 4), either in computer-based laboratory experiments or real-world field experiments using both target-present and target-absent lineups. The results showed that replication is a better technique than concealment. Compared to concealment, replication increases target identifications in target present lineups—in some cases by decreasing foil identifications in target-absent lineups. The hybrid similarity (HS) model of face recognition was used to assess whether it could be applied in this domain. Across seven experiments (Studies 1, 2, and 3) and three paradigms, the HS model was able to model the qualitative pattern of results.The purpose of this experimental work was to demonstrate the importance of constructing fair lineups for people with distinctive features and to provide results that will have practical implications for legal contexts and will improve our understanding of face recognition and recognition memory in general."--Abstract.