Brainstorming "Doesn't Work" Article Mis-interprets Smith/Kohn Study

Jena McGregor (@jenamcgregor) wrote an interesting piece recently for the Washington Post titled “Why Brainstorming Doesn’t Work.”

Jena McGregor, Washington Post

Smith of Texas A&M

Her piece interprets some rather exciting new creativity research done by Steve Smith of Texas A&M and Nicolas Kohn of University of Texas as saying brainstorming doesn’t work.

That’s not what the study is saying really.

The research by Smith and Kohn is valid and helpful in my view — and does indeed illustrate one of the challenges  of group brainstorming. Essentially, Smith and Kohn say that when people see ideas others have generated in a group brainstorming session the ideas influence — actually inhibit — the groups ability to generate a diverse set of ideas. The test group that didn’t see each others ideas and worked independently had a “broader” set of ideas. Fair enough. And…

Note that both groups were un-facilitated, not trained (given a few guidelines only) and used computers for data entry. They were also not a “team” as many corporate groups are. I’d also add “minimal motivation” — these were psychology students asked how they might improve their university.

The results of the study are not surprising. Many in the innovation services field and business leaders have seen this thing happen. Call it herd mentality, or box-creation, it’s true in some groups that the thinking around you influences and may limit your thinking. Good on Smith and Kohn for providing us hard data on this tendency.

Now, I resonate with the general tone of McGregor’s article, she describes a scene we’ve all been part of: “…with flip charts at the front of the room and candy on the table. The all-hands emergency meetings to come up with ideas to fix the latest mess. … the offsites in drab hotel ballrooms…” These kinds of sessions are indeed set up for failure, and yes, we’ve all been there.

But this study doesn’t mean brainstorming doesn’t work.  It means that a particular kind of brainstorming works better than another kind of brainstorming in a very restricted context. This new research suggests to me that in any brainstorming session there should be multiple modes of ideation, and multiple tools and techniques used to generate ideas.

A good brainstorming session (one designed to be effective) mixes people up in different ways through the session. Some idea generation should be done alone, some in pairs, some in larger groups. Some should be spoken out loud, other bits should be done as a silent writing exercise. A well-designed brainstorming session should have kinesthetic, visual, musical, and emotional components. Stimuli also can change the direction and output of a group brainstorming session. Using computers directly impacts the tone of a session (it would be interesting to run Smith/Kohn’s study without them). Preparation helps, as brainstorming participants have more interesting and diverse output when they simply have more time to incubate. Training and practice in brainstorming is also a huge factor in idea generation fluidity. Further a well planned and organized brainstorming session should be facilitated — this per the man who invented brainstorming, Alex Osborn. There was no group facilitation in the Smith/Kohn study, nor done in the 1958 study (Taylor, Berry, Block) McGregor cites as proof that individuals are more creative alone than in groups. The research of Sid Parnes, and others show different results than Taylor, et al.

Saying that brainstorming doesn’t work based on the Smith/Kohn study is simply over-reaching the limited focus of the study. What makes for effective group brainstorming research would have to include about 30 other factors to be realistic as a true test of  why brainstorming works or doesn’t work. I don’t fault Smith and Kohn, they’ve provided us with new data about one important factor.

I’ve ranted about this before and apologies to my regular readers, but I have to try once again to get people to define brainstorming. If you think of brainstorming as a one dimensional event, where a group of un-facilitated, unprepared people try to jam ideas about something they have no emotional investment in, I’d have to admit, McGregor is right in her article title. However, if you define brainstorming as I do (and as many well trained people in this field do, and how Osborn originally defined it) it would include a whole lot more than that — and one would have to conclude the title is hype and the article’s analysis is faulty.

    • Kind of upsetting how “ok” the data is here..they didn’t really do any type of “brainstorming”..not surprised this happens every three months at all.

      They should have talked to you!

    • Jeffrey Phillips also wrote a lovely piece about this Reuben, check out:

      Many people think of brainstorming as that half-baked thing they’ve experienced…so therefore the McGregor article will feel right, it confirms their experience. Reuben you’ve had the opportunity to be trained by a good teacher and have read some of the literature, including Jack’s Notebook, which makes you more of an expert than 99% of the people who say it doesn’t work.

    • The article is available online for those who want to look into it in detail. Gregg caught the basic issues reported in the article. There are a few more. On the positive side, it is a good reminder to facilitators that quantity may not bring quantity if the group is fixated in one category of ideas or there are other group dynamics that are taking place. The article describes them well. The main problem with this type of research is that it does not replicate well what happens in reality (lack of ecological validity).

      • Yes, not what happens in real life. Brainstorming is a holistic experience and must be evaluated that way. Thanks for your comment Hector. Hope you and Suzanna are well.

    • Agree with what you say, Gregg. Also, these studies also ignore what happens in the real world of business. Let’s say people go away, do solo idea generation then come back with their output. Then a group brainstorms and come up with ideas of similar, or even lesser, quality. Which idea is most likely to make it to market, “Jean’s idea” or “our idea”? Of course I’m asking these questions rhetorically because shared ownership is the best launch to life an idea can have. It also ignores the fact that ideas per se are valueless, they only generate value in implementation, and if they don’t get to market they remain useless.

    • tim


      Good on ya. I’m tired of this headline as well. I’m with you. The information in the article is great. I remember Steve from his presentation at Crea Univ at Rene Descartes University in Paris.. the same year you did a keynote

    • Pingback: Gregg Fraley | Ideas Aren’t Innovation()

Posted in Creativity and Self-Expression, Idea Generation, Innovation