Two test questions from Michael Lewis’s The Undoing Project (184-5, 192-3).
Q. In a family of six children, which birth order is more likely? BBBGGG or GBBGBG?
A. Most say the latter, but they are equally probable, equally random. People have a mental model of “randomness,” and the second matches it better than the first.
Q. One group was asked to estimate the product of 8 x 7 x 6 x 5 x 4 x 3 x 2 x 1. Another group was asked to estimate the product of 1 x 2 x 3 x 4 x 5 x 6 x 7 x 8.
A. The first group’s median answer was 2,250; the second group’s was 512. The answer is 40,320. Using 8 as a starting point tricked people into thinking the number was big (though not as big as it actually was). Starting with 1 tricked them into estimating small. Kahneman and Tversky calls this “anchoring”: Our guesstimates are affected by starting points, even when they are not relevant to the question at hand.