At airport security checkpoints and passport control desks around the world, security officers make split-second decisions about whether a face in front of them matches the photograph in an official document—and they're wrong far more often than most people realize. Now researchers at Adelaide University and the University of Stirling have uncovered a remarkably simple way to fix the problem: asking someone to take another look.
The finding matters because unfamiliar face matching is everywhere. Whether it's verifying a driver's license, checking a passport, or confirming an identity at a border, people make these decisions dozens of times daily. Yet the accuracy is sobering—laboratory studies consistently show that the average person makes mistakes in about 10 to 30 percent of face-matching decisions, roughly one error for every five comparisons. When you scale that across millions of identity verifications happening globally, even modest improvements in accuracy have real consequences.
Dr. Daniel Carragher and his team at Adelaide University conducted an elegant study published in the Quarterly Journal of Experimental Psychology. They asked participants to view 80 pairs of unfamiliar faces on three separate occasions and judge whether each pair showed the same person or two different people. The results revealed something counterintuitive: participants achieved around 80 percent accuracy on each individual attempt, but they didn't always reach the same conclusion on the same face pairs. Some pairs they got right once and wrong twice. Others they got wrong once and right twice.
When researchers combined just two of the same person's decisions, accuracy improved by around 6 percent. Combining all three attempts improved it by around 8 percent—a gain remarkably comparable to what specialists achieve through dedicated face-matching training programs, despite participants receiving no training or feedback whatsoever. "What changed was the decisions themselves," Dr. Carragher explained. On particularly challenging pairs—photos that looked nearly identical but showed different people, or very different-looking photos of the same person—participants didn't always judge the same way twice. But they were more likely to be right two times out of three than wrong two times out of three, so combining their responses filtered out one-off errors.
The research builds on the well-established "wisdom of the crowd" effect, the principle that combining many people's judgments typically produces better outcomes than trusting a single person. What's remarkable here is that a similar effect can occur within a single person reviewing the same decision again. Combining decisions from different people still produced larger gains—around 8.5 percent with two people, up to 12 percent with three—but the single-person improvement is significant given how simple it is to implement.
Dr. Carragher emphasized that more research is needed before real-world recommendations can be made. "We need to better understand how this effect translates outside the laboratory," he said. But the implication is clear: in situations where accuracy matters most, there may be genuine value in the oldest piece of advice: take a second look.