The researchers had some predictors watch a video of the experiencer’s facial expressions. They had others see only the picture the experiencer was observing, and a third group saw both the video of the experiencer as well as the picture he or she was observing.
The predictors who saw the pictures experiencers were rating were dramatically more accurate than those who only saw the video of the experiencer’s expressions—more accurate 92 percent of the time in one experiment. Watching the video of experiencers’ expressions did not seem to do anything to increase accuracy. People who saw both the video and the picture the other person was rating were no more accurate than those who simply saw the picture. Once predictors had put themselves in the role of experiencers, by looking at a picture, “reading” the experiencer didn’t further improve accuracy.
But people aren’t aware of this discrepancy. In a follow-up study, participants tended to overestimate the importance of being able to read someone’s expressions and to underestimate the importance of being in his or her situation. Participants believed that seeing a video of another person’s expressions would make them more accurate, when accuracy actually came from knowing what the other person was observing.
The findings suggest that people have too much confidence in their ability to read people. If you truly want to understand someone else, sharing his or her experience is much more effective.