Is there a name for the bias of comparing numbers on a relative scale?

Here are a few examples.

The car salesman sells you a \$900 navigation system as an up-sell to a \$30,000 car. It doesn't seem like that much because it 900 is a relatively small percentage of 30000. And let's say a top quality system installed, away from the dealership is \$700; a difference of \$200.

What causes us to think \$200 isn't that much?

But if shopping in a store where the top priced GPS was \$700 and all the others are priced around \$500, now \$200 seems too much.

Or how about a homeowner trying to sell a house for \$200,000 when a potential buyer bids at \$185,000 because of water damage in one room. The seller accepts even though the cost to repair the damage might be only \$10000. The \$5000 doesn't seem like that big of a loss for the hassle of having to do the repair, but in a lot of other circumstances (like selling a car) it's a deal breaker.

Another. Take a trip to a neighboring town and you might go a mile or two out of your way to see an attraction you just discovered. Take a trip across the country and you might visit a whole other state, hundreds of miles out of your way.

I know that anchoring bias can cause you to, for instance, perceive a price as being lower because of false higher price. But is it the same as this relative scaling bias? Or is this called something else altogether?

  • $\begingroup$ This has to do with the "utility" of these differences. The higher the values are (I.e. the more money you have), the larger the losses/gains should become for it to "matter". Look for Kahneman and Tversky on utility and value. Bet you'll get a more elaborate explanation there :) $\endgroup$ Sep 8, 2016 at 7:47

1 Answer 1


It's been called “relative thinking” in a few places (eg http://journal.sjdm.org/11/10921/jdm10921.html). The earliest I can find is in a 2004 article by Ofer Azar. A Google Scholar search turns up many other references to the term as well.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.