When Crowds Collide

Posted · Add Comment

CTBT Intensive Policy CourseCrowdsourcing has become one of the buzz words in the collaboration and innovation lexicon. It seems useful when focusing a lot of minds on solving a problem or finding a particular piece of information. Yet when looking for opinions, reactions, and input to a product, it raises an important question: When is a crowd not a crowd?

According to recent cross-cultural studies by Dr. David Rand, a lecturer in psychology at Harvard University, and Dr. Benedikt Herrmann of Nottingham University in the U.K., a crowd is a crowd only when the people’s backgrounds and the crowd’s composition are relatively homogeneous.

To study the homogeneity of crowds, Dr. Rand used what has become known as the trolleyology experiment, a classic thought experiment in which subjects are faced with a life-and-death drama. The dilemma is simple — a trolley car with no brakes will ram into and kill a group of people crossing the tracks unless the person in the experiment chooses to throw a single individual under the trolley car to slow it down enough for the others to escape unharmed. Kill one to save many, or do nothing and everyone dies. A classically simple dilemma with an incredibly hard solution.

Using crowdsourcing as the basis for the experiment, the answers from one crowd to another should have been the same. Yet the answers depended on which crowd was chosen to participate. The answers from one crowd were very different than the answers from another. In the work done by Dr. Herrmann, the results from a crowd in the United States were very different than the answers from crowds in Asia, Eastern Europe, or a Middle Eastern crowd.

Why is this important? Crowdsourcing is becoming SOP (standard operating procedure) for many companies that want customer input to innovate or co-create new products. Crowdsourcing in this instance is often viewed as a focus-group on steroids. The difference is that focus groups are carefully selected and rigorously questioned. It is as much art as science. Crowds are, well, crowds, especially when the invitation goes out on an open social network.

The experiments conducted by Dr. Rand and others conclusively show that different cultures more than any other variable provide different crowds. Human behavior is not universal. The key is to know the source before you look for the crowd.

Source: “The Roar of the Crowd,” The Economist, 05/26/12