Eyewitness suspect identification: six claims regarding the state of the science

Psychological science on eyewitness suspect identification has a long and rich history. A few decades ago, modal expert opinion emphasized eyewitnesses’ fallibility, and it was widely held that identifications made with high confidence are only slightly more likely to be accurate than those made with low confidence. The authors of this invited Contemporary Discussion agree that current science compels a more nuanced perspective in which the relationship between eyewitnesses’ confidence and their accuracy varies predictably depending on specifics of how the suspect was selected, how the identification test was designed, when and how it was administered, and when confidence was assessed. We tender claims regarding conditions under which we believe lineup identification responses can be strongly inculpating. We also articulate claims regarding conditions under which we believe identification responses can be strongly exculpating. While most of the claims described herein were previously advanced by individual scientists, what is new—and important—is that they now reflect an emerging scientific consensus. We do not assert that every claim is firmly established, but we advance arguments for believing they are true. In addition, we propose multiple lines of laboratory and field studies aimed at advancing understanding of these issues.