Reducing errors of eyewitness identifications

Kneller, W. (2002) Reducing errors of eyewitness identifications. University of Southampton, School of Social Sciences, Doctoral Thesis.

Record type: Thesis (Doctoral)

Abstract

Throughout the thesis, the issue of whether eyewitness decision strategy (relative or absolute) is indicative of accuracy was examined. Results were mixed but where an effect emerged (studies 1, 2a & 2b), the data suggested that eyewitnesses were more likely to be accurate when using an absolute decision strategy. The likelihood of using such a strategy was maximised when lineup members were viewed under a sequential presentation rather than a simultaneous presentation. In this respect, the sequential lineup was recommended over the more usual simultaneous lineup.
The effect of lineup member similarity on accuracy and decision strategy usage was also examined through the manipulation of lineup construction methods (similarity-to-suspect; match-to-description). Results indicated that neither construction method significantly influenced accuracy or decision strategy usage.
The thesis also examined the influence of an emergent factor - facial distinctiveness - with the expectation that a distinctive target would be easier to identify from a line-up because of (i) a more resilient memory trace, (ii) a representation which elicits less confusion with similar faces, or (iii) a more heterogeneous set of foils as a match to the distinctive target face. In a series of experiments (studies 2a, 2b, & 3) in which target presence, lineup construction, and lineup presentation type (simultaneous; sequential) were manipulated, distinctiveness failed to influence eyewitness performance. A partial replication of experiments of 2a and 2b in which delay was shortened also failed to find the expected distinctiveness advantage (study 4). The lack of a distinctiveness advantage was unlikely to be due to the particular targets used as a traditional laboratory-based old/new task yielded the expected advantage when recognising distinctive faces over typical faces in general, and the distinctive target over the typical target in particular (study 5).

This record has no associated files available for download.