Presenting eyewitness confidence to jury members : the influence of verbal, numerical, and graphical expressions of confidence

Eyewitness identification is susceptible to error and, despite robust evidence that eyewitness confidence can provide an index of identification accuracy, there remains uncertainty as to how confidence should be collected and presented to jurors. Current guidelines recommend collecting confidence in the witness's own words‚ÄövÑvp, but non-standardised expressions may be difficult for jurors to interpret. Two studies investigated the influence of verbal, numerical, and graphical expressions of eyewitness confidence on jury decision-making. Study One asked 44 (male=18, female=26) jury eligible participants (i.e., 18 years of age or older) to convert 16 verbal expressions of confidence to a percentage score, in terms of estimated eyewitness accuracy. There was substantial variability in participants' responses, suggesting that verbal expressions of confidence may have limited diagnostic value. In Study Two, 163 jury eligible participants read a mock police report and court transcript including testimony from an eyewitness, who identified the suspect and provided a numerical confidence estimate. Study two tested whether numerical confidence estimates facilitated framing effects that might make uncertainty salient and reduce jurors' perceptions of identification reliability. We found little evidence of framing effects for numerical expressions of confidence. However, we also found that participant did not discriminate between moderate and high confidence estimates.