Pryke, S., Lindsay, R. C. L., Dysart, J. E., & Dupuis, P. (2004). Multiple independent identification decisions: A method of calibrating eyewitness identifications. Journal of Applied Psychology, 89, 73-84.
Past research has shown a few main points of information that are relevant to the current study. First, Wells and Lindsay (1980) introduced the concept of a diagnosticity ratio in relation to eyewitness memory. This ratio is indicative of the degree to which eyewitness identifications are useful in assessing guilt of a suspect. Second, past research on the effects of sequential lineups and the effects of these lineups on correct and false identification rates were discussed. Finally, a study by Lindsay et al. (1987) was brought up to point of the methodology of multiple lineups presented to witnesses that each contain different type of information (i.e. face lineups and clothing lineups).
Experiment One
The main purpose of experiment was to test the potential advantage of multiple lineups that featured different features of the subject (i.e. face, voice and body). Each lineup contained only these features.
Method
Participants viewed the target live during a one minute speech asking for experiment volunteers. The target then left the room and the participants viewed lineups. The lineup for face and body were viewed simultaneously, while the voice lineup was completed sequentially. Some of the participants viewed target-present lineups, and the rest viewed target-absent lineups.
Results
See Table 1 for overall diagnostics
The diagnostic scores suggest that multiple target identifications across the
lineups increase accuracy, the current break down of scores cannot be tested through inferential statistics. However, a recombination of the data will allow some testing.
See Table 2 for suspect choice diagnostics
A Chi-Square suggests these diagnostics do significantly deviate from the chance expectations, even when individuals who did not choose a face are excluded.
Experiment Two
Since past research has suggested sequential lineups as a better means of presenting a lineup, the second experiment conducted all lineups in this manner.
Method
The second experiment was conducted in the lab, with a mock theft committed by the target. The lineups were presented sequentially and there were four lineups in this second experiment (i.e., face, voice, body, and clothing).
Results
Table 3 for Overall Diagnostics
Table 4 for Suspect Choice Diagnostics
The results of experiment one were
replicated. Additionally the addition of
a fourth lineup appears to have improved diagnostic ratings. Importantly, this happened in the context of
sequential lineups.
General Discussion
While the presentation of multiple lineup data may be detrimental to the prosecution’s case in the courtroom, the false identification rate under the condition of multiple identifications (3%) is lower than what has previously been found in single identification studies (9% - sequential and 27% - simultaneous). Additionally, this type of information may be helpful in working toward the suggestions of the Devlin commission, which suggested that courts use some type of evidence corroboration, as opposed to a single eyewitness identification. This type of multiple identification testing may provide another avenue to reach that goal.