Download files
Abstract
Psychologists, legal practitioners and scholars share the knowledge that honest eyewitnesses can err in their attempts to identify the perpetrator of a crime. This thesis reports an experimental investigation of the extent to which expert evidence and judicial instruction can improve juror ability to discriminate between accurate and inaccurate identifications. Special attention is also paid to the logic of inferences which have been made by psychologists regarding the efficacy of expert evidence, and compares methodologies adopting direct measures of participant Sensitivity to Eyewitness Accuracy (SEA) with those that can only indirectly assess this construct.
Study 1 surveys the knowledge and opinions of legal professionals regarding eyewitness identification issues (n = 35), showing that respondents expressed doubts that judicial instructions would exert an effect equivalent to that of eyewitness expert evidence. Accordingly, Experiments 1 to 4 (Experiment 1, n = 104; Experiment 2, n = 238; Experiment 3, n = 228; Experiment 4, n = 297) were conducted to directly assess the relative impacts of judicial instruction and expert evidence on participant juror SEA.
The methodology utilised in these investigations incorporated the testimony of real eyewitnesses to a staged crime scenario in order to assess the impact of instruction on juror ability to discriminate between known accurate and known inaccurate eyewitnesses. Overall, little evidence was found to support the notion that expert evidence is more effective than judicial instruction, as no significant association was identified between instruction type and SEA. This result was found to hold irrespective of the objective quality of the expert?s testimony (accurate or erroneous).
In light of the results from Experiments 1 to 4, Experiment 5 was designed to investigate why the experts were not able to improve the discrimination accuracy of the jurors. This study focused on the extent to which participants of varying levels of expertise could correctly classify eyewitness accuracy. The results of Experiment 5 (n = 145) suggest that experts were no better able to discriminate between accurate and inaccurate eyewitnesses than novice laypeople. Overall, the evidence reported in this thesis raises serious questions regarding the utility of eyewitness expertise in the completion of eyewitness discrimination tasks.