Don't Trust the Applause

Cheers. Applause may have less to do with the presentation and more to do with the audience.


The next time you hear extended applause for a performance you didn't think was that great, don't feel like a snob. A new study reveals that audience response has more to do with the people in the seats than those up on stage.

Applause is a bit like a pandemic, according to previous research. It begins with a few individuals and then catches on with more and more people until everyone is "infected." And as the applause dies down, the pandemic disappears. But researchers didn't fully understand the dynamics of what happens in the audience.

Enter Richard Mann, a mathematician at Uppsala University in Sweden. He and colleagues filmed groups of 13 to 20 students at the University of Leeds who were clapping after a number of different oral presentations by undergraduate and postgraduate students. Then the team took the videos back to the lab and analyzed them.

Applause, it turns out, is a bit like peer pressure. Individuals were more likely to start clapping if a larger percentage of the audience had already started, Mann's group reports online today in the Journal of the Royal Society Interface. If 50% of the audience was clapping, for example, individuals were 10 times more likely to start clapping than if 5% of the audience was clapping. People stop clapping for the same reason.

Mann and colleagues made a couple of other interesting observations. People decided to start or stop clapping not based on what they saw, but rather on what they heard. You don't have to witness the people clapping to become infected, Mann says; you just have to hear them.

Even more surprising, the applause for a bad presentation could be just as long as applause for a good one. "You should be careful about interpreting how good a talk is depending on how long the applause was," Mann says. Random interactions in the audience can result in very different lengths of applause regardless of the quality of the talk.

Other studies have shown that high-risk actions like booing can depend on a tipping point: A certain number of people have to start doing it before the momentum picks up and the rate at which people join in increases. There was no similar threshold with applause, Mann says. Even though individuals were more likely to clap when they heard others doing so, there was no minimum number of people clapping that suddenly caused a huge surge in participants.

Jon Kleinberg, a computer scientist at Cornell University, applauds the study. "It's an interesting way of measuring a kind of human dynamic process that takes place over a short window of time," he says. Kleinberg also notes that applause is a unique social behavior to study because people are immediately aware of what everyone else around them is doing. So remember that the next time you put your hands together.

Follow News from Science

dancing shoes
Volcano erupting at night