Think Twice: Harnessing the Power of Counterintuition by Michael J. Mauboussin

Author:Michael J. Mauboussin
Language: eng
Format: mobi
ISBN: 9781422176757
Publisher: Harvard Business Review Press
Published: 2012-11-05T14:00:00+00:00

FIGURE 6-1

High-dimension contests increase the uncertainty of outcomes

Correlation and Causality—Not!

Stock-market soothsayers are always looking for reliable ways to anticipate the market’s direction. One favorite is the Super Bowl Indicator, invariably trotted out after the football season’s championship game. The indicator is simple: the stock market goes up when a National Football Conference team wins and goes down when an American Football Conference team wins. The Super Bowl winner has correctly predicted the stock market’s direction nearly 80 percent of the time from 1967 to 2008. Another is David Leinweber’s analysis that shows a 75 percent correlation between butter production in Bangladesh and the level of the Standard & Poor’s 500 Stock Index (1981–1993). Leinweber mined a wide range of international data series and was pleased to find that “a simple dairy product” explained so much.15

Leinweber used a silly example to make a serious point: the failure to distinguish between correlation and causality. This problem arises when researchers observe a correlation between two variables and assume that one caused the other. Once you are attuned to this mistake, you will see and hear it everywhere—especially in the media. Vegetarians have higher IQs. Nightlights lead to nearsightedness. Kids who watch too much television tend to be obese.

Numerous scholars from varied disciplines have studied causation, and most agree that three conditions must hold to make a claim that X causes Y.16 The first is that X must occur before Y. The second is a functional relationship between X and Y, including the requirement that cause and effect take on two or more values. For example, the statement “smoking causes lung cancer” says that smoking increases the chances of lung cancer versus not smoking. So a scientist must consider all the relationships between the variables: does the person smoke (yes or no) and does the person have cancer (yes or no). Here, too, you must consider whether the relationship is merely happenstance.

The final condition is that for X to cause Y, there cannot be a factor Z that causes both X and Y. For instance, watching too much television may correlate with obesity. But low socioeconomic status may explain both the television viewing and the weight problem.17

You must be very alert to the correlation-causality mistake. The fact that we like to make explicit cause-and-effect connections only adds to the challenge. When you hear of a causal connection, step carefully through the three conditions to see if the claim holds up. You will most likely be surprised at how rarely you can firmly establish causation.

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.