Picking someone out of a lineup can be tough. Bias in eyewitness testimony is a complicated factor that can influence judicial proceedings. When eyewitnesses identify a criminal, they need to recognize a person’s face, and they need to place them at the scene of the crime. That’s a difficult task for most passersby. We might think someone looks familiar, but it takes extra brain power to remember exactly where and when we saw them. It becomes particularly challenging in cases that have a lot of media attention, or in small towns where “everyone knows everyone.”
Many studies in eyewitness memory research assume that the observer only saw the guilty person at the crime scene. But what about cases where the observer also saw the guilty person before the crime occurred? This happens more than you might think. Eyewitnesses report prior familiarity with a criminal in anywhere between 10% and 67% of cases (Valentine et al., 2003; Flowe et al., 2011; Bruer et al., 2017).
Researchers have found that prior familiarity can affect face recognition in two ways:
- It helps with a correct identification: “Yup, I know that guy, and he definitely robbed that bank.”
- It can lead to a false identification: “I think that man robbed the bank… or maybe he was a customer at the bakery.”
One limitation to some studies of “familiarity” is that researchers train participants on one facial image. During testing, they use that same image to ask if participants saw that face before. The problem with this is that it does not generalize well to real-world scenarios. When we see people in the world, we rarely see them in the exact same position, lighting, background, etc. Therefore, a more practical application would include multiple images of the target faces.
In a recent article in Psychonomic Bulletin & Review, Belgin Ünal, Melisa Akan, and Aaron S. Benjamin used various images of faces to test recognition of “criminals” after different levels of familiarity.
The study used a three-phase procedure:
- Familiarization Phase: faces of “neighbors” were viewed, and some had names attached
- Study Phase: faces of “criminals” were viewed; some of them appeared in the familiarization phase, and some did not
- Recognition Phase: participants were shown faces and needed to decide if the person was guilty (criminal) or innocent (non-criminal)
When it came to the recognition phase, participants were given six options to make their assessment: “Sure Guilty,” “Maybe Guilty,” “Guess Guilty,” “Guess Innocent,” “Maybe Innocent,” and “Sure Innocent.”
The researchers measured hit rates—or the times when guilty faces were correctly identified as guilty—and false alarm (FA) rates—or the times when an innocent face was incorrectly assumed guilty.
The results show that conceptual processing—assigning a name to the face—helps with correctly identifying a guilty person. On the flip side, any familiarization leads to higher rates of guilty responses, regardless of whether the person is guilty or innocent. That is, familiarity can increase correct hits AND false alarms, but adding the conceptual knowledge of a person’s name helps discriminate the truly guilty faces.
The authors suggest that conceptual familiarization may be superior to perceptual familiarization in enhancing memory. They conducted this study in an attempt to create greater real-world application compared to previous studies. The information gained may be highly relevant for cases that involve eyewitness testimony, and it will certainly be on my mind the next time I listen to a true crime podcast.
Featured Psychonomic Society article:
Ünal, B., Akan, M., & Benjamin, A. S. (2024). Prior familiarity enhances recognition memory of faces, not just images of faces, when accompanied by conceptual information. Psychonomic Bulletin & Review, https://doi.org/10.3758/s13423-024-02576-3
