Gut Feelings Contains Deep Lessons For Journalism and Publishing

gerd instinct

08Oct07

Gerd Gigerenzer’s new book, Gut Feelings: The Intelligence of the Unconscious (Allen Lane, £14.99) has been troubling me for some time. It’s the kind of book whose title, like The Wisdom of Crowds, is going to mislead a few people. There are already a few too many out there who worship their own gut feelings, and no amount of education or reason will shift them. Put them in a group, and…well, you understand why the Daily Mail is such a successful newspaper.

To be sure, though, the Director of the Center for Adaptive Behaviour and Cognition at the Max Planck Institute has not written a defense of ignorance. If anything, the opposite. He’s in the business of explaining how some of these instincts, when applied to the correct problems, actually do yield the best results.

When a risk of collision exists

Simple rules of thumb (heuristics) may encapsulate more complex calculations, and this could save your life. The so-called gaze heuristic, which we use when we catch a ball, is also taught in inverse to pilots and sailors to avoid collision. I spent many a cold winter evening in night school navigation classes learning that if an object is coming toward you on a constant bearing, you can bet that it’s going to hit you. [Don’t go sailing with anybody who struggles with that concept.] The alternative would be to calculate one’s real-time position in four-dimensional space. This is one of those instances where, in fact, the best advice would be “don’t do the math”.

Chest pains and a doctor’s gut instincts

Fear of litigation frequently leads US clinicians to “play it safe” when a patient presents with chest pains. Patients are sent to intensive care; there, if they’re not having a heart attack, they are exposed to a risk of infection that could be equally injurious. Twenty thousand Americans die each year from infections contracted in hospital. That’s a gut reaction that’s not good.

In a Michigan hospital, the result of this defensive decision-making meant overcrowding in intensive care, deteriorating quality of care, and rising costs. Gigerenzer shows how the situation was first improved by a complex method of calculation provided by a team of medical researchers from the University of Michigan. But when the calculator was withdrawn, and doctors had to rely on their intuitions again, their diagnostic performance did not drop back.

What seems to have happened is that the doctors internalised the cues from the calculator and improved their intuitions to the point that they did not need to run the numbers, which they did not like doing anyway. But there is a further problem; risk factors may differ from hospital to hospital and area to area, so creating the complex calculation for each environment would be too costly.

By using what Gigerenzer describes as a fast and frugal decision tree, a process of elimination of different cues can be quickly worked through that highlights the most relevant diagnostic steps, simply and effectively. This is one of the key points that Gut Feelings aims to reveal; that in complex situations, where there is a high degree of uncertainty (heart attack prediction being notoriously difficult) simple diagnostic methods tend to be more accurate than more complex ones.

The fast and frugal tree is much less information-hungry than a traditional decision tree and can be characterised by three factors:-

Search Rule: Look up factors in order of importance.
Stopping Rule: Stop the search if a factor allows it (i.e. take the best answer when it first presents itself).
Decision Rule: Classify the object according to this factor. Or, or in other words, come to a diagnosis.

Generous to a default

Gigerenzer also highlights how the framing of a choice can lead to different outcomes depending on the default option. For example, there was a lot of hot air circulating in the UK recently about proposals to change the rules concerning organ donation from the current one of opting in to one of presumed consent. In France, donation rates are much higher than in the UK and US, because the default presumes consent. Very few people subsequently opt out.

This was quite useful as it reminded me to update my own status. I realised that the cardboard donor card that had once been in my wallet must have completely disintegrated at some point in the past 20 years, leaving my altruistic aspirations completely unrecognizable. (By the way, you can do it all online now, here for the UK, and here for the US.)

The small-scale furore that blew up in the UK insisted that the presumed consent approach was “un-British”. The fact that more than 400 people die every year while waiting for a kidney, lung, heart or liver transplant would not seem terribly “British” either. In the US, 19 people die each day waiting for an organ transplant.

Gigerenzer highlights that such decisions work on the basis of inertia, and that if policy is set for an opt-in, few people actually will decide to act against the majority/default. Gigerenzer calls it the default rule. We apply it thus without even thinking about it:-

If there is a default, do nothing about it.”

Collaboration, coffee and cake

Gigerenzer has some subtle advice for managers. He shows how the attitudes of leaders can create implicit behaviours throughout an organization and that these can endure for a very long time — long after a manager has departed. For example, a leader’s aversion toward unnecessary email can induce a reticence to communicate upwards; an excessive concern for presenteeism can discourage time out of the office at conferences where new ideas or intelligence might be gathered. He therefore urges executives to think carefully about what values their unwritten rules communicate.

Gigerenzer describes how he himself wanted to build a more collaborative culture at the Max Planck Institute:-

Unless one actively creates an environment that supports this goal, collaboration tends to fall apart within a few years or may never get off the ground in the first place. The major obstacle is a mental one. Researchers, like most ordinary people, tend to identify with their ingroup and ignore or even look down on neighboring disciplines.”

Gigerenzer instituted a number of rules which he says have endured long after he stopped organizing them — like daily gatherings around coffee. To reinforce a cultural imperative that all success was shared, anyone who had something to celebrate, like an award or an acceptance of a paper for publication, was obliged to bring cake to the coffee meetings.

Anyone who has ever sought to sustain collaboration within a workplace will admire Gigerenzer’s insistence that when the institute expanded, the researchers were kept on the same floor by extending the building laterally. He says it was his experience that employees interact 50 per cent less when occupying different floors. The loss of productivity will rarely be shown on an accountant’s spreadsheet. Of course, the option is not always available if you are squeezed — as I was for a decade — into an Edwardian building in Fleet Street.

Related Posts

wall street journal in peril cries wolf

how “chippy” do you like your ice-cream?

bloomberg not real-time at all