Liquor Stores and Churches

They say there are lies, damn lies, and statistics. Here’s an example of why this is sadly true in many cases.

It’s been proven that there is a strong correlation between the number of churches and liquor stores in a community.

Some conclude then that the key to keeping the number of liquor stores down is to stop allowing all those churches to go up! If all you did was look at the graph of these two variables it would be hard to argue the fact.

But the key to unlocking this mystery is to understand the difference between correlation and causation. Two things can be correlated strongly, liquor stores and churches for example, without causation. 

This is to say that even though there seems to be some relationship between two variables we need to stand back and ask if the results make sense.

In the church and liquor store example further probing would lead you to learn that the growing population of the community is the true reason there are more churches, liquor stores, hair dressers, donut shops, etc.

So the next time some fancy statistician tries to sell you on how two things are correlated just remember they may be right but before turning down a nice preacher wanting to build a church do your homework and confirm there is causation.

For more details on this correlation, causation, and even a little extrapolation fun check this post out.

Lastly, no disrespect meant to my Baptist friends with the picture… but it was too good to pass!  Credit to the picture goes to this site.

If you enjoyed this post please subscribe to this blog via RSS feed.

2 Comments

  1. Gary Petersen

    September 29, 2007 - 1:37 pm

    But I like the statistics. Back in the day, when I was in Six Sigma black belt training, the statistician teaching that section used data of increased consumption of ice cream and increased rate of violent crime in Miami. Not likely a cause and effect relationship there, which was exactly his (and your) point.