Business Research = Groupthink

I don’t really know what groupthink means – it sounds Orwellian. If it is then it means what it sounds like (or the opposite). Anyway, I define ‘Groupthink’ as a pervasive and gradual development of a highly constrained and conforming view of the world, that is grounded in the hope that crowds do have wisdom. Of course, recent events pretty much demonstrate that crowds are dumb.

I recently read a so called research paper in which a particular and currently popular view of the world was stated, the supported by interviews with a large number of ‘executives’ who were asked what they thought would be their biggest challenges in the next couple of years. These kinds of studies are very common in the IT world, and in the business literature as well. The survey responses are then collated, run through the statistics program and presented in graphs. Then the author provides a few helpful ideas, backed up by case studies, on how we might all plan for the new world order.

There are many reasons why this is wrong. As a scientist I am often confronted by people asserting stuff without a shred of evidence (from my view anyway). In this instance, in order to demonstrate the truth of an hypothesis, a survey was taken of many people’s opinions. This is anecdotal evidence, and valuable for helping a researcher to design an experiment to test it. But it is not data. A common aphorism in science is ‘the plural of anecdote is NOT data.’ It doesn’t matter that you measure it with a scale that gives you numbers that you can manipulate in MS Excel. It doesn’t matter that you can print out pie graphs that say ‘ 30% of executives think that x will happen in the next 2 years.’ It is still just what people reckon will happen.

But worse than this, providing helpful ideas backed up by case studies to address situations that might or might not happen is dangerous. Businesses are very complex entities and just because Dow corning were able to achieve a particular result with a particular approach doesn’t mean it will work here. Where are the case studies of where the approach did not work?

For me the problem is not that the surveys are done, or reported or anything like that. It is that the way the information is presented lends it a level of ‘scientific’ credibility that it is not due.

So the end result is that many readers of business magazines read about surveys of what CEOs worry about – they then start to worry about those same things, and hey presto – it all comes to pass.

I started into this to get something off my chest – and I sort of have – but for those interested in a more thoughtful critique, get hold of a book, Hard Facts by Pfeffer and Sutton.
Here is a video of them talking about their book.