Regression Analysis > Spurious Correlation
What is a Spurious Correlation?
A spurious correlation wrongly implies a cause and effect between two variables. For example, the number of astronauts dying in spacecraft is directly correlated to seatbelt use in cars:
This graph — showing that an increase in car seat belt results in a lower number of astronaut deaths– is made from real statistics. For example, the 1980s seat belt use data came from this journal article on PubMed. Of course there isn’t a real correlation here: putting your seat belt on in a car has nothing to do with the odds of an accident in space. This kind of connection which seems to be real, but isn’t, is called a spurious correlation.
The sad truth is that spurious correlations are everywhere: in the news we read, on blog posts and news sites, and all over the airwaves. Take this gem from Fox News, showing a clear correlation between president Obama taking office and a rise in unemployment rates:
Look closely at the graph, and you’ll see it’s plotted incorrectly. The value for November (8.6) is placed at the 9.0 position. That’s not the only problem with this graph; it’s been plotted to show a steep incline when in fact, the overall employment trend was fairly stable at about 9%.
Even if the data you’re looking at is sound, there may be other factors causing the phenomenon. For example, take the headline Sunny Skies tied to Suicide Rates. While short-term exposure to sunny skies may contribute to suicide rates, there’s obviously a lot more to suicide than a depressing, sunny day.
Examples of Spurious Correlation in the Media
Spurious correlations don’t always have graphs attached to them. They are often a result of a media frenzy, misunderstanding of data, or just plain old bad science.
Universal health care breeds terrorists (Fox News). An on screen headline on Fox News read “National healthcare: breeding ground for terror?”. I’m not sure what the statistics are for the number of people who signed up for Obamacare and then turned into a terrorist, but I’m willing to bet it’s pretty small. Fox News didn’t use any actual data for this spurious correlation, just the opinion of an “expert.”
Living Next to Freeways Causes Autism (L.A. Times). Hot on the heels of the vaccine-causes-autism debate (which, by the way, has been debunked) we find that people living next to freeways are at a higher risk of having an autistic child. But before you think about the correlation too hard, consider the confusing statement released by the paper’s lead author, Heather Volk: “This study isn’t saying exposure to air pollution or exposure to traffic causes autism..but it could be one of the factors that are contributing to its increase.” In other words, a possible correlation has been turned into a definite link by way of a misleading headline.
Junk food does not cause obesity. The non-profit Global Energy Balance Network reported that consumption of junk food isn’t to blame for the obesity epidemic. The solution to losing weight, said the group, was simply to exercise more. Before you think you can eat all of the junk food you like and not put on weight, consider who sponsored the group: one of the world’s great junk food producers, Coca Cola.
As a final note, be aware that several studies have suggested that watching Fox News makes you stupid. A spurious correlation? You be the judge.
Next: Misleading Graphs.
If you prefer an online interactive environment to learn R and statistics, this free R Tutorial by Datacamp is a great way to get started. If you're are somewhat comfortable with R and are interested in going deeper into Statistics, try this Statistics with R track.Comments are now closed for this post. Need help or want to post a correction? Please post a comment on our Facebook page and I'll do my best to help!