Face-matching -- employed at places like airports, where TSA and customs agents check IDs and passports -- is not a foolproof system, a new study shows.
In fact, the error rate is anywhere from 10 to 20 percent in a lab setting -- which means that it would likely be even higher in the real world, noted researchers from Louisiana State University and Arizona State University.
"Because society relies on face perception and ID verification for many tasks, people are often under the impression that we are experts in this domain," study researcher Megan Papesh, an assistant professor at Louisiana State University, said in a statement. "Our research shows the precise opposite."
The study, which is published in the journal Attention, Perception and Psychophysics, included four different experiments. For the experiments, researchers had study participants look at pairs of photographs of faces that were taken 1.5 years apart. The researchers wanted to test to see how the prevalence of identity mismatches -- in other words, the prevalence of having two faces not actually be the same person -- affected the ability of the participants to detect the mismatches. The researchers did this by comparing the ability to detect mismatches when just 10 percent of the photograph pairs were identity mismatches (a low prevalence), and when 50 percent of the photograph pairs were mismatches (high prevalence).
Researchers found that the participants were more likely to miss an identity mismatch when the prevalence of mismatches was low, missing 20 percent of the mismatches when the prevalence was low but missing more than 40 percent of the mismatches when the prevalence was high. "This effect persisted when participants were allowed to correct their initial responses (Experiment 2), when they had to verify every decision with a certainty judgment (Experiment 3) and when they were permitted 'second looks' at face pairs (Experiment 4)," the researchers wrote in the study.
The findings may apply especially to situations like face-matching passports with travelers, where the rate of fake or stolen IDs may be low but the risks of allowing a user of a fake/stolen ID to travel may be high.
"Unfortunately, these conditions are also those most likely to give rise to poor detection rates," Papesh said in the statement. "In our research, when observers infrequently encountered fake IDs, they failed to catch approximately 45 percent of them, even when given multiple opportunities to correct their errors."