In the story of Pinocchio, his nose grows when he lies. Real noses do not grow. But scientists in Spain used special cameras to see heat. They found that the nose often got colder when people told a lie.

Other studies found that the nose got warmer. So the results are mixed. At an airport, human interviewers found lies better than the cameras did.

Researchers at the University of Granada in Spain used thermal cameras, which show heat as colors, to study what they called the "Pinocchio effect." In a 2017 study, people told an important lie to someone close to them on the phone, and the temperature of their nose tip and hands tended to drop.

However, in a 2016 Italian study, the nose became warmer in some situations. The Granada team later explained that stress, mental effort and other factors can push the temperature in different directions. In a 2011 airport study, thermal cameras correctly classified about two-thirds of travelers, but human interviewers did even better. Another problem is that honest people may also feel nervous and show the same changes.

Can a thermal camera catch a liar? Researchers at the University of Granada used the term "Pinocchio effect" after observing temperature changes around the nose during deception. In their 2017 study, participants phoned someone close to them and told a significant lie. The tip of the nose and the hands tended to cool, which the team attributed to stress-related changes in blood flow.

Other results pointed the other way. A 2016 Italian experiment, in which participants freely chose whether to lie for personal gain, found that the nose warmed in some conditions. In 2018 the Granada group itself concluded that emotional tasks tend to cool the nose while cognitively demanding tasks warm it, so the direction depends on the situation. The Granada experiments were laboratory tasks, so real-world accuracy is uncertain.

Anxious but honest people may also show identical responses, which makes false alarms a serious concern. Field tests have not been encouraging either. In a 2011 study at an international airport, thermal imaging correctly classified roughly two-thirds of 51 passengers. Human interviewers did better, and giving them the thermal data on top of their own impressions did not improve their accuracy.