Confidence inflation in eyewitnesses: Seeing is not believing


Correspondence should be addressed to Amy Bradfield Douglass, Department of Psychology, Bates College, 4 Andrews Road, Maine 04240, USA (e-mail: ).


Purpose. Confidence inflation in eyewitnesses obscures a useful cue to identification accuracy and affects evaluations of eyewitnesses (e.g., Bradfield & McQuiston, 2004; Jones, Williams, & Brewer, 2008). We examine whether sensitivity to confidence inflation evidence is enhanced by seeing a videotape of the identification procedure.

Methods. Participants (N= 131) watched a videotaped trial in which the witness's original confidence statement was presented as part of a previously recorded videotaped identification procedure or read by the witness at trial. In addition, the witness's identification confidence was either consistently high or low at the time of the identification and high at the trial (i.e., it was inflated).

Results. Significant interactions demonstrated that confidence inflation evidence factored into judgments of the eyewitness and defendant guilt more strongly in the videotape condition compared with the read condition.

Conclusions. The present results support recommendations to collect immediate confidence reports and videotape identification procedures. Using videotape evidence may help innocent defendants convince jurors that the eyewitness's identification is not accurate.