7.2.2. How It Works
In the previous calculations, anchors people tend to use are higher or lower depending on the first digit of the multiplication (which we read left to right). The anchors then unduly influence the estimate people make of the answer to the calculation. We start with a higher anchor for the first series than for the second. When psychologists carried out an experimental test of these two questions, the average estimate for the first series was 4200, compared to only 500 for the second.
Both estimates are well below the correct answer. Because the series as a whole is made up of small numbers, the anchor in both cases is relatively low, which biases the estimate most people make to far below the true answer.
In fact, you can give people an anchor that has nothing to do with the task you’ve set for them, and it still biases their reasoning. Try this experiment, which is discussed in Edward Russo and Paul Schoemaker’s book Decision Traps.2
Find someonepreferably not a history majorand ask her for the last three digits of her phone number. Add 400 to this number then ask “Do you think Attila the Hun was defeated in Europe before or after X,” where X is the year you got by the addition of 400 to the telephone number. Don’t say whether she got it right (the correct answer is A.D. 451) and then ask “In what year would you guess Attila the Hun was defeated?” The answers you get will vary depending on the initial figure you gave, even though it is based on something completely irrelevant to the questionher own phone number!
When Russo and Schoemaker performed this experiment on a group of 500 Cornell University MBA students, they found that the number derived from the phone digits acted as a strong anchor, biasing the placing of the year of Attila the Hun’s defeat. The difference between the highest and lowest anchors corresponded to a difference in the average estimate of more than 300 years.
Taken from : Mind Hacks
