I feel the need to interrupt my day-by-day recap of my Texas-Louisiana vacation, in order to talk about an interesting team-building exercise I participated in today at a work-related event.
I usually make it a policy not to talk about my job on this blog, at all. Because, you know, you can get fired for that. However, since I’m not revealing any confidential information, or disparaging anyone, I think I can make an exception here.
The exercise was as follows. Each participant was given a sheet of paper with 12 multiple-choice questions about wilderness survival (i.e., “Given scenario X, would it be best to do a, b or c?”). After answering the questions individually, everyone was asked to form groups of 6, and then use a “consensus-building” approach to come up with group answers to the questions. (“Consensus-building” means that you have to get everyone in the group to agree to the group answer before moving on; you can’t just have a vote and put down the most popular answer.)
Most of the groups scored better on the exercise than most individuals did. And, with the exception of one group (if I’m remembering this correctly), no one individual had a higher score than their group. It seems to me that the obvious lesson has to do with the value of a consensus-building approach to problem-solving.
However. Another interesting observation was that no one group had a higher score than the highest-scoring individual within that group (EDIT–See note at bottom). Also (this is the part I hope I’m remembering correctly), in the lowest-scoring group, the individuals tended to have higher scores than the group score arrived at through consensus-building.
I’m not an expert on these sorts of studies. But it seems to me that, at least in this exercise, while a majority of people did better on the test by working on it in teams, the teams didn’t appear to gain anything from the contributions of the less-knowledgeable team members. And, in the lowest-scoring team (presumably the one that coincidentally ended up with fewer knowledgeable people), most people ended up doing worse as a result of pooling their knowledge.
I wouldn’t conclude from this exercise that a consensus-building approach is bad. But I might suggest that the reason it’s effective (when it is effective), is that the people on the team are able to recognize who the “experts”* are, and defer to them. And I might further suggest that, if the team doesn’t have any experts, there isn’t much to be gained by pooling their expertise (rocket science, I know).
Donald suggested that one lesson of this exercise might be that a dictatorial leadership style is more effective than a consensus-building approach. Assuming that your dictator is the most knowledgeable person in the group, of course. I don’t think I agree. Or, at least, I don’t entirely agree. The thing is, most real-life situations, at least in scientific research, are more complicated than the fairly simplistic wilderness survival scenarios we were presented with (each with only 3 possible answers). So I think it’s less clear, among scientists, who the expert is. On cross-disciplinary projects, a single person is even less likely to be the expert in every aspect of the project.
So, hmm, maybe I did come away with the right lesson after all?
I still maintain, though, that (A) you have to have experts, and (B) people have to recognize when someone is more knowledgeable in a certain area, and defer to their expertise.
Another thing I noticed: With questions that I wasn’t absolutely 100% sure about, the group convinced me to switch from a right answer to a wrong one just as often as they convinced me to switch from a wrong answer to a right one.
In any case, an interesting exercise.
* I use the term “expert” loosely, when refererring to the success enjoyed by members of our department in identifying the best wilderness survival strategies. The highest score was 8 out of 12. (“Dammit Jim, I’m a chemist, not a Sherpa!”)
(My score was 7 out of 12, in case you’re wondering. Group and individual.)
(EDIT–Okay, I remembered only after I wrote this that there was actually one team where the group score was higher than any individual scores. I don’t know how much higher it was. So perhaps synergy can occasionally work. There was only one team out of 5 or 6, though, where this was the case. I suppose the lesson here is that it’s easy to forget about data that doesn’t appear to support one’s hastily drawn conclusions.)