KARL MOORE – This is Karl Moore of the Desautels Faculty of Management at McGill University, talking management for The Globe and Mail. Today, I am delighted to speak to [Columbia University professor] Sheen S. Levine.
You have been looking at studying interactions and found something, I think, quite surprising. Tell us more about that please.
SHEEN LEVINE – So Karl, people interact constantly. They share knowledge in organizations, they get together in conferences, they chit-chat in cocktail parties, and now we have more opportunities to interact than ever, thanks to video conferencing and telecommuting and social media.
But how does interaction affect our skills, our performance, and our success? It is often assumed and is taken as an article of faith that the more we interact, the more we share, the more we talk, and the more we brainstorm, the better we will do. If you read John Stuart Mill, he speaks about the value of putting people in the company of others that are different from themselves, and this is supposed to elevate their performance. The idea is that when we interact, exchange opinions, good ideas will rise and bad ideas will dissipate.
The problem is there is hardly evidence that this is what happens and what we found in our research is that interaction not only improves your performance, but interaction can harm your performance. In fact, I would say, on average, when we interact, we often hear more bad advice than good advice.
KARL MOORE – This seems counterintuitive. Why do you think this is?
SHEEN LEVINE – When we look at people's true skills, and when we look at people's self-assessment of their skills, we find a disparity. The disparity is such that the worst performers are the most overconfident.
So if we look at people's performance on a scale from zero to 100, those that perform at the bottom, say 12.5 per cent out of 100, believed that they performed at around 60 per cent. So there is a big gap between how well they performed and how well they think they performed. When we look at the high performers, say people who performed at 85 per cent out of 100, we find the opposite – these people actually underestimated their performance and were underconfident.
So in our experimental simulations, we look at a way skills change with interactions, and we find that when you interact with other people, when you interact widely, when you go to conferences and chit-chat in the hall, and go to cocktail parties, your performance does not really improve but it just becomes very similar to everyone else.
So some poor performers do learn, and improve, but some top performers deteriorate because they hear the wrong ideas and mistakenly adopt them.