Correlation and Causation – The Case of the Glazed Donut Diet

by Sally Boone
(Detroit, USA)

Does eating a hot, glazed donut every Friday cause weight loss? Let’s say that a man named Joe has been steadily losing weight for a few months. His weight loss started at about the same time he made a habit of visiting the local donut shop every week. Is Joe onto a new diet fad?

The answer of course is a resounding no, of course. But circumstances like Joe’s pop up all the time. When two things happen in the same time frame, people oftentimes assume that they are directly related and maybe one even caused the other. But it takes a closer look at all of the circumstances to determine the real causes and effects.

There are many times when a person might want to analyze the relationship between two variables. For instance, a salesperson wants to know what happens to a store’s profits when new coupons are offered, or a scientist wants to know how a pilot’s blood pressure changes when a rocket accelerates. In instances like these, the salesperson or scientist wants to know the correlation between the variables Does more money come in when coupon are available or not? Does blood pressure go up, go down, or not change at all as speed increases?

Association is a term used to describe how two variables are related and is often interchangeable with the word correlation. At this point, no one is saying that one thing causes another. Taking a look at the correlation is merely an attempt to see if there is some type of relationship or not. Variables can have positive or negative associations. Don’t get confused, though! This has nothing to do with associations being good or bad.

A positive association is when both variables go up or down together. For instance, maybe the scientist found that as a rocket accelerates, the pilot’s blood pressure goes up. A negative association is when the variables go in opposite directions, such as if the salesperson discovered that after handing out a lot of coupons, the store’s profits went down.

Joe’s weight was coming off at a steady pace, but he was also chowing down on a fresh donut every Friday. There is a positive correlation there. However, there is a popular phrase that addresses this correlation doesn’t imply causation. In other words, the association between donut eating and weight loss doesn’t automatically mean eating the donut causes the weight loss.

Determining actual causation comes from closely studying an association between variables while also understanding that there are other possible variables and explanations out there. Some cases can be simple. A hammer smashing a thumb causes pain. There’s not a lot of mystery there. But simple cases like that tend to make people think that all cases of apparent association also indicate causation.

Maybe Joe is on a strict diet, which is causing the weight loss But the diet makes him so hungry that he breaks down every Friday and has a donut. The single donut isn’t enough to counteract his overall diet, so his weight continues to go down. The donuts didn’t cause the weight loss But the weight loss-or more specifically, the diet – did lead to the donut eating.

It’s important to not jump to conclusions when you’re presented with two seemingly related things. Taking the time to look into the circumstances more thoroughly can save you from future headaches or embarrassment. A donut weight loss diet might sound great, but it won’t take long for reality to prove that correlation doesn’t imply causation.

Please Follow & Share:
Updated: August 31, 2013 — 4:43 pm

Site Disclaimer: This site is designed for educational purposes only and is not engaged in rendering medical advice or professional services.
If you feel that you have a health problem, you should seek the advice of your Physician or health care Practitioner.

Frontier Theme