Repeated postevent questioning can lead to elevated levels of eyewitness confidence.

Abstract

Two experiments tested the proposition that postevent questioning can lead to later increases in witness confidence without corresponding changes in witness accuracy. After a staged interruption in a college classroom, participants were questioned about the event 5 times over 5 weeks in Experiment 1 (n = 57) and 3 times over 5 days in Experiment 2 (n = 79). During the final questioning session, the participant-witnesses consistently reported higher levels of confidence for those items that had been subject to repeated postevent questioning than for those items that were asked for the first time, yet there was no difference in the accuracy of the responses to the two sets of items. Additionally, in all conditions the participant-witnesses were generally overconfident in their responses. These results suggest that repeated postevent questioning can cause eyewitnesses' subsequent confidence estimates to be "artificially" inflated.

The Curious Complexity between Confidence and Accuracy in Reports from Memory

“…Shaw found that such exposure greatly infl ated subjects' confi dence in their answers but had no eff ect on their accuracy. In a related study, Shaw and McClure (1996) showed that repeated questioning infl uenced witnesses' confi dence without increasing their accuracy. Th is eff ect was also replicated by Odinot et al (2009;but see Ebbesen & Rienick, 1998, and a review by Knutsson, Allwood, & Johansson, 2011, that shows that the issue is still in some doubt).…”

“…Assuming that Shaw and McClure (1996) and Wells and Bradfi eld (1999) are correct, then unlike manipulations that aff ect accuracy and then induce a correlation with confi dence, one can manipulate confi dence without accuracy showing a corresponding increase. Obviously, to the extent that this latter fi nding occurs, problems can be created for the criminal justice system.…”

“…Th us, our recommendation that confi dence should be taken into account applies most strongly to police interrogations. Repeated questioning of witnesses has sometimes been found to increase confi dence without increasing accuracy (Shaw & McClure, 1996). By the time a witness arrives in court (oft en months or even years aft er the occurrence of the crime), confi dence may be relatively fi xed by prior tests.…”

Cheap talk and credibility: The consequences of confidence and accuracy on advisor credibility and persuasiveness

“…In other words, confidence can be an important source of information (Bonaccio & Dalal, 2006;Yaniv, 1997). Confidence and accuracy often co-vary (Sniezek, 1992) but the relationship can be weak (Deffenbacher, 1980;Kassin, 1985;Klayman, Soll, González-Vallejo, & Barlas, 1999;Shaw & McClure, 1996;Wells, Lindsay, & Ferguson, 1979;Wells, 1993) and sometimes they are uncorrelated (Brewer & Wells, 2006;Ekman & O'Sullivan, 1991).…”

What do we really know about witness preparation?

“…However, objectively false but subjectively true testimony can be created when a witness's memory of an event is distorted during the course of witness preparation, leading them to give unknowingly false or misleading testimony. 4 Memory distortion can occur when suggestive questioning by attorneys or trial consultants causes witnesses to recall events incorrectly (see Loftus, 1975;Loftus & Hoffman, 1989), and when repeated questioning (e.g., during rehearsals) causes initially uncertain witnesses to become certain (see Shaw & McClure, 1996;Turtle & Yuille, 1994). In both circumstances, the content of the testimony is changed by the witness preparation procedure.…”