Who can recognize unfamiliar faces? Individual differences and observer consistency in person identification.

It can be remarkably difficult to determine whether two photographs of unfamiliar faces depict the same person or two different people. This fallibility is well established in the face perception and eyewitness domain, but most of this research has focused on the "average" observer by measuring mean performance across groups of participants. This study deviated from this convention to provide a detailed description of individual differences and observer consistency in unfamiliar face identification by assessing performance repeatedly, across a 3-day (Experiment 1) and a 5-day period (Experiment 2). Both experiments reveal considerable variation between but also within observers. This variation is such that the same observers frequently made different identification decisions to the same faces on different days (Experiment 1). And when new faces were shown on each day, observers that produced perfect accuracy on one day made many misidentifications on another (Experiment 2). However, a few individuals also performed with consistent high accuracy in these tests. These findings suggest that accuracy and consistency are separable indices of face-matching ability, and both measures are necessary to provide a precise index of a person's face processing skill. We discuss whether these measures could provide the basis for a selection tool for occupations that depend on accurate person identification.