William Poundstone writing in his book Priceless:
Daniel Kahneman and Amos Tversky “used one piece of apparatus, a carnival-style wheel of fortune marked with numbers up to 100. A group of university students watched as the wheel was spun to select a random number. You can play along – imagine that the wheel is spinning right now and the number is… 65. Now answer this two-part question:
(a) Is the percentage of African nations in the United Nations higher or lower than 65 (the number that just came up on the wheel)?
(b) What is the percentage of African nations in the United Nations?
Like many experiments, and some wheels of fortune, this one was rigged. The wheel was designed to produce one of only two numbers, 10 and 65. This rigging was done only to simplify analysis of the results. In any event, Tversky and Kahneman found that the allegedly random number affected the answers to the second question. The effect was huge.
When the wheel stopped on 10, the average estimates of the proportion of African nations in the UN was 25 percent. But when the wheel of fortune number was 65, the average guess was 45 percent. The latter estimate was almost twice the first. The only difference was that the estimates had been exposed to a different ‘random’ number that they knew to be meaningless.
Tversky and Kahneman used the term ‘anchoring and adjustment’ for this. In there now classic 1974 Science article, ‘Judgement Under Uncertainty: Heuristics and Biases,’ they theorised that an initial value (the ‘anchor’) serves as a mental benchmark or starting point for estimating an unknown quantity. Here, the wheel of fortune number was the anchor. The first part of the question that the subjects compare the anchor to the quantity to be estimated. Tversky believed that the subject then mentally adjusted the anchor upward or downward to arrive at their answers to the second part of the question. This adjustment was usually inadequate. The answer ended up being closer to the anchor than it should be. To someone inspecting only the final outcomes, it’s as if the anchor exerts a magnetic attraction, pulling estimates closer to itself.
By the way, how did your answer compare to the 65-group’s average of 45 percent? In case you’re wondering, the correct fraction of African UN member nations is currently 23 percent.”
Nassim Nicholas Taleb summarises this experiment and provides some further examples in his book The Black Swan:
“Kahneman and Tversky had their subjects spin a wheel of fortune. The subjects first looked at the number on the wheel, which they knew was random, then they were asked to estimate the number of African countries in the United Nations. Those who had a low number on the wheel estimated a low number of African nations; those with a high number produced a high estimate.
Similarly, ask someone to provide you with the last four digits of his Social Security number. Then ask him to estimate the number of dentists in Manhattan. You will find that by making him aware of the four digit number, you will ellicit an estimate that is correlated with it.
We use reference points in our heads … and start building beliefs around them because less mental effort is needed to compare them with to a reference point than to evaluate it in the absolute.”