Over the last couple of months, I have had several people tell me that research has shown that brainstorming does not work and they generally refer to a blog by Matthew Syed, “Please, no more brainstorming sessions. This is not how innovation really works” (https://www.linkedin.com/pulse/please-more-brainstorm-sessions-how-innovation-really-matthew-syed), or a New Yorker article by Jonah Lehrer, “Groupthink. The brainstorming myth” (http://www.newyorker.com/magazine/2012/01/30/groupthink). Both of these refer to a research by Charlan Nemeth (See end of blog for details). The two bloggers claim that Nemeth’s paper shows that allowing debate in an idea generation session give 25% more ideas and, if this is true, then brainstorming, where debate is not allowed, is a waste of time. Unfortunately, Nemeth’s paper does not show this. While the groups encouraged to debate ideas generated on average 21.7 ideas per group and groups given traditional brainstorming instructions generated 18.7 ideas per group, these differences were not statistically significant. In other words, no difference was found between Brainstorming and Debate. This prompted me to do some reading into the effectiveness or otherwise of brainstorming. The most interesting of the papers I came across are listed at the end of this blog. Based on these published works, I have come to the following conclusions.
1) Most of the research into brainstorming has indicated that it is no more effective, and often less effective, than a “nominal group”. The nominal group technique involves a number of people generating ideas individually and then aggregating their ideas.
2) Most of the research into brainstorming uses small groups, generally around 4 or 5 people.
3) Most of the research into brainstorming uses psychology students, without any experience of brainstorming, as test subjects.
4) Most of the research into brainstorming uses groups without a facilitator trained in brainstorming.
5) Most research into brainstorming looks at idea generation in isolation, using model (i.e. not real) problems that are of limited interest to the participants.
On this basis, what the research shows is that, small groups of psychology students, without a trained facilitator and working on a problem of limited interest to them, produce fewer ideas than the total generated by the same number of people working independently. It may be that this can be extrapolated to the population as a whole but, in that much use phrase, more work is required to confirm this.
(In the interest of full disclosure, I have only carried out a quick Google and JSTOR search and so may have missed many papers that counter some of these conclusions. If anyone knows of any peer reviewed articles looking at the effectiveness of brainstorming I would be very interested to hear of them.)
Of the five conclusions, possibly the most important is the last. When we are trying to solve a problem, the final output does not come from an isolated ideation session, whether it uses brainstorming, SCAMPER, analogies, or any of the hundreds of techniques that have been promoted. Ideation is not even the most important step in solving a problem. Before we start looking for ideas we need to spend some time studying the problem so that we fully understand what it is we are trying to do and what any potential solutions need to deliver. We can then think about generating ideas for potential solutions.
After we have some ideas for potential solutions we need to pick out the ones that are most likely to deliver a robust solution. Often, the hard work carried out in an ideation session is wasted because this step is rushed. If we have 100 ideas, it is unlikely that we will be able to try out all of these so we need to reduce them to a more manageable number (around 10 to15). Often this is done by “dot voting”, but this can lead to the rejection of the more off-the-wall ideas in favour of the ideas that fit the group’s comfort zone. I have found that it is better to reduce the number gradually using two or three filters focused on different aspects of the problem.
In the ideation step, the ideas will usually be presented as short phrase on a Post-it or moderator card. This is what I call bullet-point thinking. In order to implement a potential solution, we need to build more detail into these ideas. To do this, we focus on one idea and challenge it. Why would we want to do it? What will prevent us from doing? How do we change it to make it more do-able without losing the benefits we will get? Focusing repeatedly on each of these questions will build the idea into something that is more viable. We can then move onto the next idea.
Finally, once we have concepts for potential solutions, we need to prepare them for a decision on which to implement. Since implementation will rely on people who were not involved in the generation of the concept, we need to record it in a form that they will understand. The details required will depend upon the problem but there are some generic questions we should answer – what exactly are we proposing, why do we think we are capable of doing this and what benefit will we get by doing it? In doing this, we need move away from the bullet point thinking that we have used up until now and use more “joined-up” thinking. The more detail we can get the better. I have found that, in workshops, it is better to do this stage with groups of 3 people, rather than the groups of 7-9 that would be used in ideation.
There are many “thinking tools” available for each of these steps. I run occasional open seminars looking at a selection of these (including brainstorming in the form of Brainwriting), and showing how they can build together to form an effective framework for problem solving or new product idea generation. See the Perth Innovation website for details of upcoming seminars in Aberdeen and Perth.
Does Group Participation When Using Brainstorming Facilitate or Inhibit Creative Thinking?
Donald W Taylor, Paul C Berry and Clifford H Block
Administrative Science Quarterly, Vol3, No 1, (Jun 1958), pp 23-47
Brainstorming Groups in Context: Effectiveness in a Product Design Firm
Robert I Sutton and Andrew Hargadon
Administrative Science Quarterly, Vol 41, (1996), pp685-718
The Liberating Role of Conflict in Group Creativity: A Study in Two Countries
Charlan J Nemeth, Bernard Personnaz, Marie Personnaz and Jack A Goncalo
European Journal of Social Psychology Vol 34, (2004), pp365-374
Collaborative Fixation: Effects of Others’ Ideas on Brainstorming
Nicholas W Kohn and Steven M Smith
Applied Cognitive Psychology, Vol 25, (2011), pp359-371.
PMI Aberdeen EventBlinking Fast and Slow (The importance of cognitive ...
There are stories, going back to ancient Greece, of dolphins helping stranded ...
Her: So, what do you do?Me: I help companies to think out of the box and act in ...
In my work, I have frequent discussions with people about innovation. These ...