The Iowa gambling task: How to improve your hunches with data
Intuition has long gotten a bad rap in analytics circles. Search for “gut-feel business intelligence” in Google, and you’ll likely find a myriad of articles that disparage intuition and champion rational analysis as the more effective strategy.
Clearly, one reason for the bad attitude towards intuition is the self-serving interest of analytics vendors. Another is that analytics really are superior to gut feel in many situations. But more recently – as highlighted in the following books and many others – there’s been a resurgence of focus on intuition and the fact that it can sometimes be an appropriate strategy:
- In Thinking Fast and Slow, Daniel Kahneman shines a spotlight on the two different systems of thought and the advantages of each
- Gary Klein, in Streetlights and Shadows, makes a case for blending intuition and analysis to capture the best of both worlds
- In How We Decide, Jonah Lehrer recounts how real-world decisions rely on a mix of feelings and reason
Books like these and related articles often refer to a seminal research method first used more than 10 years ago. It started as a clever way to measure decision-making strategies in individuals and evolved into a second version with even deeper implications for intuition in business.
The Iowa gambling task – original recipe
In 1994, Iowa-based neuroscientist Antonio Damasio and his team of researchers published their first paper describing a “novel task which simulates real-life decision-making in the way it factors uncertainty of premises and outcomes, as well as reward and punishment.”
In the method, test subjects were given a “loan” of $2,000 in fake U.S. bills and asked to play a game with four decks of cards. The objective: maximize profit on the loan.
They could select one card at a time from any of the four decks in return for a sum of money. Sometimes, the subjects would also have to pay a penalty that varied in size. They weren’t told in advance how long the game would last, but it always ran until 100 cards had been selected. Unknown to the subjects, the four decks were rigged:
- Decks A and B always paid out $100 per card but had an average penalty of $125 per card
- Decks C and D always paid out $50 per card but had an average penalty of $25 per card
By counting the number of selections subjects made from each deck, the study measured how quickly they learned that decks A and B were bad and should be avoided. Not surprisingly, they found that most subjects picked far fewer cards from decks A and B in the last 50 selections than they did in the first 50.
The Iowa gambling task – hunchified
In 1997, Damasio and his team kicked things up a notch in a new study. Could they gauge when a person’s emotional response to the bad deck of cards actually began?
To do this, they ran the same gambling task as before, but this time hooked up electrodes to the hands of the subjects. The electrodes were there to measure skin conductance and help quantify the emotional reaction to a stimulus.
In the study’s widely cited results, Damasio’s team identified several stages of cognition:
- The pre-hunch stage, roughly between the 10th and 50th cards, when subjects didn’t suspect that there might be a bad deck
- The hunch stage, between approximately cards 50 and 80, when they “reported ‘liking’ or ‘disliking’ certain decks…but were not sure of their answers”
- The conceptual stage, after about the 80th card, when subjects could accurately articulate their task and knew for certain which were the good and bad decks
What’s interesting, though, was what the electrodes showed. In the pre-hunch period, just before a subject selected a card from a bad deck, the electrodes measured a significantly higher emotional response. In other words, their emotions flagged the deck as bad even before they began to consciously suspect something was amiss.
Even more surprising was that during the pre-hunch period, subjects also selected fewer cards from decks A and B. It’s as if their gut was subconsciously telling them to avoid the bad decks, despite the fact they had no conscious knowledge that their card-selection strategy was changing.
Lessons for business
The original goal of Damasio’s studies was to compare control subjects with others who had damage to their prefrontal cortex and see how their responses differed. But his results can also be instructive in a business context – as lessons for how to improve decision making.
Build quick feedback into your processes
Damasio and his team created a system with immediate feedback. When a subject selected a card, he’d win money or pay a penalty. He didn’t have to wait until the next day, week, or quarter to link cause and effect. It’s a no-brainer to extend that lesson to your organization – the faster you report the results of your actions, the better.
Use data to train your hunches
Once you have a system that provides immediate and frequent feedback, use it to develop your intuition. Whether it’s sales targets, marketing numbers, or production metrics, if you can accurately link your action to an emotional response (triggered, perhaps, simply by viewing positive or negative data), you may start making better decisions even before you can verbalize your hunches. At the very least, you’ll have a head start on others who insist on slower, traditional decision-making strategies.
Eugene Sadler-Smith, a U.K.-based researcher on organizational behavior, goes even further in connecting Damasio’s results and scores of related studies to explicit lessons for executives. In his paper, The Intuitive Executive, he suggests several ways to harness and build intuition in a business setting.
Open your decision-making closet
The key, says Sadler-Smith, is to admit that intuition is “automatic and involuntary,” especially if your natural tendency is to suppress your gut feelings before you’ve had a chance to vet them. In truth, experienced leaders often admit they prefer intuition, but if intuition is suppressed, it may stop working. No one learns if executives hide their intuitions.
Ask for constructive feedback
Build an environment that encourages constructive criticism. Find someone “to give honest, accurate, and constructive feedback on [your] behaviors, attitudes, and performance.” If you surround yourself with staff who don’t challenge your decisions, you’ll learn the wrong lessons and hamper the ability to refine your intuition accurately.
Seek out a devil’s advocate
Don’t just ask for general feedback – actively seek out someone who can “raise objections to favored choices, challenge underlying assumptions, and point out alternatives.” This process can expose weaknesses in your intuition and help make it stronger in the long run.
A magic bullet?
Is there a single magic bullet – one tool or process – that alone will help you grow your organization’s intuitive strengths? Of course not, but by changing your attitudes about intuition – and coupling them with tools for fast analysis and an environment that supports collaboration and feedback – you’ll be on the right track.