{"id":110609,"date":"2023-08-08T14:33:00","date_gmt":"2023-08-08T18:33:00","guid":{"rendered":"https:\/\/www.shortform.com\/blog\/?p=110609"},"modified":"2023-08-16T13:18:03","modified_gmt":"2023-08-16T17:18:03","slug":"correlation-doesnt-equal-causation","status":"publish","type":"post","link":"https:\/\/www.shortform.com\/blog\/correlation-doesnt-equal-causation\/","title":{"rendered":"Steven Pinker Explains Why Correlation Doesn\u2019t Equal Causation"},"content":{"rendered":"\n

Do you ever confuse correlation with causation? How can you accurately determine the cause of something?<\/p>\n\n\n\n

In Rationality<\/em>, Steven Pinker examines why people make irrational decisions<\/a>. Often, people run into problems when considering causation and correlation. He explains how to avoid the trap of linking them when they’re not connected and offers some tips on how to determine actual causes.<\/p>\n\n\n\n

Continue reading to understand why correlation doesn\u2019t equal causation.<\/p>\n\n\n\n\n\n\n\n

Correlation Doesn\u2019t Equal Causation<\/h2>\n\n\n\n

A common mistake people make is thinking that events that are correlated<\/em> (they often happen at the same time) are causing<\/em> each other, when in fact they might be linked simply by coincidence or by a third factor. The failure to understand that correlation doesn\u2019t equal causation can lead people to make poor decisions<\/a>; when they think the wrong event causes another, they incorrectly predict the future<\/a>.<\/p>\n\n\n\n

For example, if the stock price of a company always rises in November, a person might think the arrival of November causes the price to rise, and they might then buy stock in October in anticipation of that rise. However, if the true reason behind the price increase is that the stock rises when the company offers a huge sale on their goods, which they happen to always offer in November, then the person buying stock in October might lose money if the company decides not to offer that sale this particular November. If the person had correctly identified the causal link (between the sale and the price rise, instead of between the month and the price rise), they might have purchased their stock at a better time.<\/p>\n\n\n\n

Pinker notes that it can be difficult to determine causation, especially when there are multiple events or characteristics to account for. Complicating matters is that, very often, correlation does imply some sort<\/em> of causation: If two events are commonly linked, they likely have a common source<\/strong> (as in the stock price example above).<\/p>\n\n\n\n

How Our Search for Meaning<\/a> Misleads Us<\/strong>

In Fooled by Randomness<\/em><\/a>, Nassim Nicholas Taleb<\/a> argues that one reason people confuse correlation with causation is that humans are wired to seek meaning<\/a>, which drives us to invent meaning when we see patterns of events. This can lead us to mistake correlations that are due to either random chance or a third factor that\u2019s not as easy to detect.

This can be a particularly easy trap to fall into if the events we\u2019re studying are closely related in theme<\/a> as well as in timing, as this often indicates a common source. For example, eating ice cream and getting sunburned are both related to hot afternoons. However, if you get a sunburn every time you eat ice cream, it\u2019s not the ice cream that\u2019s causing your burn\u2014it\u2019s the common source of both events: the hot sun.<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n

When determining what factors cause other factors and which are merely correlated, Pinker notes that you can do one of two things:<\/p>\n\n\n\n