Close

June 25, 2013

Don’t (always) trust your intuition

A few month ago, we talked about how intuition can be misleading. In brief, we all tend to trust our intuition more than we should. All of us. Yes, that includes you.

Not convinced?

Here is a little exercise to see how well you are doing (again, this is stolen from Bazerman and Neale, p.56-58):

Listed below are ten requests for information you are unlikely to know. For each write down your best estimate of the answer. Then put a lower and upper bound around your estimate such that you are 95 percent confident that the correct answer falls within this “confidence range.”

1  ___ Number of General Motors automobiles produced in 1990.

2  ___ IBM’s assets in 1989.

3  ___ Total number of $5 bills in circulation on 31 March 1990.

4  ___ Total area in square miles of Lake Michigan.

5  ___ Total population of Barcelona, Spain, in 1990.

6  ___ Amount of taxes collected by the IRS in 1970.

7  ___ Average annual snowfall in Anchorage, Alaska.

8  ___ Number of bound volumes in all twenty-six branches of the San Francisco Public Library.

9  ___ Dollar value of outstanding consumer credit at the end of 1988.

10 ___ Median price of existing single family homes in Honolulu, Hawaii, in 1990.

How many of your ten ranges actually include the correct answers? If you set your ranges so that you are 95 percent confident, nine or ten of them should. The correct answers for each of these items are: (1) 3,213,752; (2) $77,734,000,000; (3) 5,772,195,480; (4) 67,900 square miles; (5) 4,163,000 people; (6) $195,722,096,497; (7) 68.5 inches; (8) 1,749,129; (9) $728,900,000,000; and (10) $290,400.

We trust our intuition more than we usually should

If you have good intuition, you should have, at most, one of the above answers wrong. But chances are that you are off, even though you could have set as wide an interval as you wanted. So your judgment is off. Welcome to the club.

This is quite a widespread problem: Arkes and Kajdasz note that confidence and accuracy generally are not closely related. They write: “An intuitive theory that predictions or judgements made with high confidence are more likely to be correct. However, research suggests that even expert decision makers such as physicians can lack a strong relation between confidence and accuracy.”

This is rather surprising and scary.

It means that you shouldn’t blindly trust your judgment. It also means that you should reduce your exposure to the risk of being wrong. In the same publication, Arkes and Kajdasz propose some ideas, pointing out that accountability and feedback are both important in improving the accuracy-confidence relation. They also reinforce the important of looking for contrary evidence when testing hypotheses, pointing out that when people do so in controlled environments, they improve drastically the confidence-accuracy relation.

If you aced the test above, you are probably safe to continue approaching problems as you do. But if you are like the rest of us (I scored a shameful 20%), take this as a wake-up call to question your judgment: unfortunately, it isn’t as impeccable as you thought it was.

References:

Arkes, H. R. and J. Kajdasz (2011). Intuitive theories of behavior. Intelligence analysis: Behavioral and social scientific foundations. B. Fischhoff and C. Chauvin, The National Academies Press: 143-168.

Bazerman, M. H. and M. A. Neale (1992). Negotiating rationally. New York, The Free Press.

Leave a Reply

Your email address will not be published. Required fields are marked *