Abstract: . . .  experimental psychology Page 1 The confidence-accuracy relationship 1 In press: Journal of Experimental Psychology : Applied Running Head: EYEWITNESS IDENTIFICATION The confidence-accuracy relationship in eyewitness identification: Effects of lineup instructions, foil similarity and target-absent base rates Neil Brewer Flinders University Gary L. Wells Iowa State University Correspondence: Neil Brewer, School of Psychology , Flinders University, GPO Box 2100, Adelaide, South Australia 5001 Telephone: 61-8-8201 2725 Fax: 61-8-8201 3877 Email: neil.brewer@flinders.edu.au . . .
. . .  experimental psychology Page 1 The confidence-accuracy relationship 1 In press: Journal of Experimental Psychology : Applied Running Head: EYEWITNESS IDENTIFICATION The confidence-accuracy relationship in eyewitness identification: Effects of lineup instructions, foil similarity and target-absent base rates Neil Brewer Flinders University Gary L. Wells Iowa State University Correspondence: Neil Brewer, School of Psychology , Flinders University, GPO Box 2100, Adelaide, South Australia 5001 Telephone: 61-8-8201 2725 Fax: 61-8-8201 3877 Email: neil.brewer@flinders.edu.au Page . . .
. . .  elicited different identification response patterns, and (b) under two independent manipulations that induced changes in identification accuracy for one or both targets. The point-biserial CA correlations, though discriminating choosers from nonchoosers, were modest across all conditions. Despite the modest CA correlations, plotting confidence against proportion correct for choosers clearly indicated a positive relationship between confidence and accuracy for both sets of stimulus materials under all experimental conditions. As shown by the point-biserial correlations, the resolution statistics indicated that confidence was useful in discriminating correct from incorrect positive identification responses. Further, diagnosticity for positive identifications made with 90-100% confidence was high. The same patterns were not evident for nonchoosers, providing the first clear evidence using an eyewitness identification (rather than a face recognition) paradigm that identification confidence and accuracy . . .
--3000,3,500,2394,58930