Using ecphoric confidence ratings to discriminate seen from unseen faces: The effects of retention interval and distinctiveness

Abstract

Theories of confidence processing for recognition judgments suggest that confidence indexes the degree of match between a presented stimulus and an image in memory (ecphoric similarity). Recent research has demonstrated that having participants rate their confidence that a face had been previously seen provides an equivalent or a better index of the stimulus's status than does eliciting a simple binary response (Sauer, Brewer, & Weber, Journal of Experimental Psychology: General 137: 528-547, 2008). Using a face recognition paradigm, we manipulated retention interval and stimulus distinctiveness to directly test the suggestion that confidence indexes ecphoric similarity and to probe the boundary conditions for using confidence ratings to discriminate seen from unseen faces. Consistent with the proposed ecphoric basis for confidence ratings, mean confidence was higher for previously seen than for unseen faces, and conditions conducive to the formation of strong memories improved confidence-based discrimination. In all conditions, after the application of a classification algorithm, confidence ratings provided a more sensitive index of face status (i.e., seen or unseen) than did binary responses.

The Relationship Between Eyewitness Confidence and Identification Accuracy: A New Synthesis

“…But research by Sauer, Brewer, and Weber (2008) found that collecting a witness confidence statement for each lineup member (rather than only the one who was chosen) provided a more informative indicator of recognition. Following on this finding, more recent research has shown promising results for procedures in which eyewitnesses do not pick someone out of a lineup at all but instead make a confidence judgment about whether each lineup member is the perpetrator (e.g., Brewer, Weber, Wootton, & Lindsay, 2012; Sauer, Brewer, & Weber, 2008; Sauer, Brewer, & Weber, 2012) or rate how well each face matches their memory of the perpetrator (Sauer, Weber, & Brewer, 2012). Results from profile analyses and classification algorithms have shown that such methods may be superior to the traditional eyewitness-identification task.…”

Mock-juror evaluations of traditional and ratings-based eyewitness identification evidence.

“…"not guilty") verdicts provided the witness demonstrated the ability to discriminate between lineup members. Sauer and colleagues (Sauer & Brewer, 2015;Sauer et al, 2008Sauer et al, , 2012aSauer et al, , 2012b have previously argued that one potential benefit of a ratings-based procedure is that ratings may indicate that a witness recognizes the suspect even in cases where a categorical response might have resulted in a lineup rejection. For example, a witness who is particularly concerned about the prospect of identifying an innocent person may reject a lineup even if they are reasonably sure that they recognize someone in the lineup.…”

Using confidence ratings to identify a target among foils.

“…This information is not available if the witness provides a single identification decision. Second, across multiple stimulus sets and recognition memory tasks, previous research demonstrates that confidence ratings provide classification performance that equals or exceeds that of traditional identification tasks, and face recognition tasks (Sauer et al, 2008(Sauer et al, , 2012. This evidence demonstrates that, when asked to provide a single identification response, witnesses do not make optimal use of the memorial information available to them.…”

“…Subsequently, classification algorithms designed to optimize classification performance determined criteria for classifying these confidence ratings as indicative of a guilty or innocent suspect (see below for further explanation of the classification procedure). Previous research demonstrated that, compared to single identification decisions, confidence ratings provide a more sensitive index of recognition (Sauer et al, 2008; Sauer, Weber, & Brewer, 2012). However, in previous research, optimum classification criteria were developed separately for each stimulus set.…”