Confounding Variables

A confounding variable is an outside influence that changes the effect of a dependent and independent variable.

It is the third wheel. It ruins the relationship.

The Ice Cream and Shark Attack Myth

There is a strong correlation between ice cream sales and shark attacks. When one goes up, the other goes up.

Does ice cream attract sharks? No.

The confounding variable is summer.

In summer, people buy ice cream. In summer, people swim in the ocean. The heat causes both. Ice cream and sharks have no relationship.

The Coffee and Cancer Scare

Years ago, studies suggested coffee caused pancreatic cancer.

It turned out that coffee drinkers were also more likely to smoke.

Smoking causes cancer. Coffee does not. But because the two were linked in the participants' habits, coffee took the blame. Smoking was the confounder.

Controlling the Chaos

Good science controls for confounders.

• Randomization: Randomly assigning people to groups distributes the confounders equally.
• Matching: Ensuring both groups have the same age, sex, and smoking status.
• Statistical Adjustment: Using math to "subtract" the effect of the confounder.

The Diagnosis

If a study claims X causes Y, ask: "What else could cause Y?"

If they didn't measure Z, they didn't prove anything.