Purpose: Research has shown that people often have difficulties estimating eyewitness accuracy correctly. In most previous studies examining validity in credibility judgments, participants have assessed the accuracy of witnesses who have been homogeneous in their memory performance. This study investigated validity in judgments of witnesses who varied widely in memory. A further purpose was to examine whether judgmental validity was moderated by the witness’ ethnic ingroup/outgroup status.
Methods: Participants (N=120) rated the credibility of videotaped testimonies of high- and low- accurate in-group (Swedish, n=4) and out-group (immigrants, n=4) witnesses who were genuinely trying to recall a criminal event.
Results: Participants assigned more credibility to high- than to low- accurate in-group witnesses, while out-group witnesses received low credibility ratings regardless of their actual memory performance. Path analyses demonstrated that the self-reported confidence of in-group, but not of out-group witnesses predicted participants’ accuracy judgments.
Conclusions: The results indicate that the validity in judgments of in-group witnesses can be better than has previously been implied. Investigators may have difficulty distinguishing high- and low-accurate witnesses from other ethnic groups, and may also systematically underestimate the reliability of ethnic out-group witnesses. Implications for legal practices are discussed.