Bias in the spotlight: anchoring

Anchoring is thought to be generated via two different effects:

• The first is through deliberate, reasoned adjustment using our rational, conscious, logical System 2. For example, when choosing wine in a restaurant we might look at the most expensive wine on the list and the cheapest and make our choice by adjusting from those prices. So, with the most expensive bottle priced at £30 and the cheapest at £10 we might choose a £25 bottle.
• The second cause is through priming effects, when we subconsciously anchor to an often unrelated number – a phenomenon that draws on our automatic, subconscious System 1, for example when bidding in an auction we might subconsciously anchor to a number we recently read, such as a bus number or phone number. If the number was high, it subconsciously drags up what we might bid, and if low it drags it down.

One piece of research even showed that the amount we repay on our credit card bill can be influenced by the relatively low number of the minimum payment in comparison to the full amount owed. People who had seen the minimum payment paid off only 23% of their balance owed, compared with those who were only shown the total balance, who paid off 40% of the balance owed.

Daniel Kahneman and Amos Tversky, pioneers of behavioural economics

To get a feel for System 1 anchoring effects, have a go at this conundrum. Try to guess, within five seconds, the value of the following arithmetical expression. 5 seconds. Ready?

1 x 2 x 3 x 4 x 5 x 6 x 7 x 8

What did you get?

In the study asking this exact same question, designed and tested by Daniel Kahneman and Amos Tversky:

• Students shown “1 x 2 x 3 x 4 x 5 x 6 x 7 x 8” made a median estimate of 512.
• Students shown “8 x 7 x 6 x 5 x 4 x 3 x 2 x 1” made a median estimate of 2,250. 

The motivating hypothesis was that students would anchor to the first few numbers then adjust upward. In case you were wondering – the right answer is 40,320. 

Crawford Hollingworth is founder of The Behavioural Architects