Penrod and Cutler

Methodology

Experiment

Mock-trial scenario

Independent measures design

Procedure

videotape of robbery trial

Witness testified either 80% or 100% sure she identified the robber

10 variables altogether

4 of the variables

Suspect in discuise

HIGH: Heavily disguised

LOW: Minimal disguise

Weapon focus

HIGH: Weapon clearly brandished

LOW: Weapon visible

Retention interval

14 days

2 days

Witness confidence

100% confident

80% confident

Participants experienced either high or low variable conditions on a random basis

Participants had to decide if robber was found guilty or not

Conclusion

Confidence is a poor predictor of witness accuracy

Jurors' trust in confidence is indiminished

Participants

Undergraduates

Eligible and experienced jurors

Results

Only significant effect was witness confidence

60% convicted when witness was 80% sure

67% convicted when witness was 100% sure