• Forum has been upgraded, all links, images, etc are as they were. Please see Official Announcements for more information

Confirmation Bias

Sure, confirmation bias refers to a poor way to argue. If research is done with a confirmation bias then it's very likely the research could reach incorrect conclusions. Say, we wanted to argue about a dog named Harvey.

Let's say that we want to argue that

Harvey is a good boy.

Is it a true statement that "Harvey is a good boy"?

Someone produces a report and argues Harvey is a good boy because:

  • Harvey fetched a stick
  • Harvey loves his people
  • Harvey gets treats for being a good boy
  • Harvey does tricks
  • Harvey does not chase cars
Therefore we conclude Harvey is a good boy.

However, last month, Harvey tore up a very expensive couch. Harvey's people have to be careful how close they let Harvey to a couch because Harvey will destroy the couch. So, we could also argue "Harvey is a bad boy."

Conformation bias happens when we only present facts that support the case, and we don't seek out or ignore facts that are contrary to the claim.

A more complete report might conclude "Harvey is often a good boy, but we should keep him away from couches."
I'm pretty sure Harvey didn't use Dash Direct to buy that couch, but who knows! :-D

First you have to ask the right question. Asking if Harvey is a good boy is not the correct question. Do you love Harvey and forgive all his imperfections? If you do, then the price of the couch and his preferred payment method becomes somewhat moot. Woof.