Person descriptions as retrieval cues: Do they really help?

Abstract

Person descriptions by witnesses are an important part of police investigations. In most cases they precede a later person identification task. While some authors believe that verbalizing faces would lead to a detriment in identification performance (verbal overshadowing), the present study focused on potential benefits of reinstatement of context by re-reading one's own person description prior to making an identification. In a laboratory experiment using a video film of a theft, re-reading one's description led to a slight improvement in identification, and a reduction in confidence ratings. Additional data on the relationship between person descriptions and identification are presented. Theoretical and practical implications are discussed.

Person descriptions and person identifications: Verbal overshadowing or recognition criterion shift?

“…While these results have been interpreted 504 SAUERLAND, HOLUB, SPORER primarily on the basis of a possibly facilitative effect of context reinstatement it should be noted that they can also be explained with Clare and Lewandowsky's (2004) criterion shift account: While the memory preserving effect due to rereading across the 1 month retention interval only occurred in the target-disguise condition also a general tendency of rereaders to not choose (55%) when compared to nonrereaders (40%) was observed. Sporer (2007) compared describers only to describers with rereading and found a nonsignificant tendency toward a facilitative effect of rereading (51.7% vs. 36.0% accuracy in the description only condition) in the expected direction. However, Sporer's experiment may have lacked statistical power to detect a context reinstatement effect.…”

“…Across k032 hypothesis tests of the relationship between description accuracy and identification accuracy with N02973 participants, the weighted mean effect size was r0.14 (pB.001), with CIs of .11 and .18. Sporer (2007) found a significant interaction between identification accuracy and choosing with accurate nonchoosers (M05.5) reporting more correct descriptors than inaccurate nonchoosers (M03.3). No effect was found for choosers (M04.2 vs. 4.6).…”

The application of multiple lineups in a field study

“…At least five methods of doing so can be found in the literature: the innocent suspect could be, first, the foil who resembles the target most according to a pilot study (e.g. Kneller, Memon, & Stevenage, 2001;Sporer, 1992Sporer, , 1993Sporer, , 2007, second, selected by experienced policemen who create the photo-spreads (e.g. Brigham et al, 1986;Fleet, Brigham, & Bothwell, 1987;Olsson & Juslin, 1999), or, third, the foil chosen most often (Pryke et al, 2004).…”

Human identification using facial comparative descriptions

“…A subset of traits from the Aberdeen University face rating schedule (FRS) [3] were used in this research. The FRS features a comprehensive selection of traits and has been used in other studies [16,6]. The FRS contains 53 absolute traits, the majority described using 5 point bipolar scales.…”

“…The FRS contains 53 absolute traits, the majority described using 5 point bipolar scales. The modified FRS introduced in [16] was used as a base for the traits used in this study.…”