Measuring the Relationship Between Eyewitness Identification Confidence and Accuracy

Abstract

The relationship between eyewitness identification confidence and accuracy has long been of interest to memory researchers. Moreover, given that (a) many police investigations and legal cases rely upon eyewitness evidence, and (b) the propensity for witnesses to err when making identification decisions, knowing whether a witness’s confidence is informative about the likely accuracy of their identification decision is also an issue of considerable forensic relevance. In this chapter we examine the different ways in which researchers have measured the confidence-accuracy relationship, showing how the method of measurement has important consequences for conclusions about the nature and strength of the relationship. We note how more recent research using calibration and confidence-accuracy characteristic approaches to measurement provide much more optimistic pictures about the strength of the confidence-accuracy relationship than the correlational approaches that long dominated measurement approaches and shaped how researchers argued that identification confidence judgments should be interpreted in the courtroom. Although data from our published studies over the last decade or so have been crucial in shaping the often very optimistic contemporary conclusions about the strength of the confidence-accuracy relationship, we also highlight a number of factors that can seriously undermine the relationship and constrain conclusions that police investigators and jurors might seek to draw about the likely accuracy of an identification.