Sensitisation Instructions Can Reduce the Misinformation Effect and Improve the Eyewitness Confidence-Accuracy Relationship

Multiple studies have reported evidence that the misinformation effect can be reduced or even eliminated under some conditions, but these studies have typically used warnings that could not be implemented in forensic settings (e.g., telling participant/witnesses that a particular source included false information). In the current study, we investigated whether novel, ecologically valid sensitisation instructions can reduce the misinformation effect. We also examined effects of the manipulation on the confidence-accuracy relationship. Across two experiments that used different stimuli and test formats, participants (total N = 422) were exposed to misinformation about a mock crime; later, half of the participants received sensitisation instructions before completing a memory test. The misinformation effect was significantly smaller for participants who received the sensitisation instructions. Sensitised participants also demonstrated a stronger confidence-accuracy relationship and were less overconfident at the highest level of confidence. Our findings encourage tests of the sensitisation instructions under more naturalistic conditions.