0

Can you give me an example for one variable being confounded with another? (I've read the tag wiki for "confounding" but I'm still confused.)

T. G.
  • 232
  • 1
  • 5
  • 1
    Does this answer your question? [Confounder - definition](https://stats.stackexchange.com/questions/59369/confounder-definition) – Adrian Keister Mar 10 '20 at 14:35

1 Answers1

0

Imagine you are testing whether people prefer product A or product B, so you make them eat product A, rate it, and then eat product B and rate it. You calculate a mean rating and see that these ratings are almost the same. So, on average, people think both products are equally as good, right? That may not be true.

A possible confounding variable here is the order in which the products were tested. If you are participating in this test, you may think product A is great and give it the maximum grade (suppose it's a 10). However, when you try product B you think it's better than product A, and because you can't rate it with more than 10 you give them the same grade, even though you don't like them the same.

I hope this example helps you understand confounding variables. I think Simpson's Paradox examples may be interesting too.

Jose Pliego
  • 371
  • 1
  • 5