T
he Conjunction Fallacy, like most supposed fallacies, drives me a little crazy. The idea is that, in certain situations, human beings mistakenly consider the occurrence of a conjunction of two independent events to be more probable than the occurrence of a single one of them, which by classical probability standards is impossible. Thus, human beings act “irrationally”.
In the most famous example, test subjects are given the following question:
Linda is a 31-year-old American woman; she is intelligent, and has a college degree; she is outspoken and cares deeply about human rights.
Which of the following statements is more likely?
1. Linda is a bank teller, and actively engaged in feminist movements.
2. Linda is a bank teller.
The majority of respondents answer that 1 is more likely than 2, despite the fact that “bank tellers who are feminists” is a sub-group of “bank tellers”, and therefore it is, classically speaking, impossible for 1 to be more likely than 2.
At first blush, this problem irks me, because it’s like crying “fallacy!” when people make an error on a trick math problem. The way of talking and asking for information implied by the question is highly technical, and divorced from the problems and methods of communication encountered by ordinary people in their ordinary lives. What is happening here is that people hear an extremely foreign type of question, and try to “read between the lines” to get at what they believe the questioner means; most of the time, their guess is not “this person wants me to affirm a statistical truism”, but “this person wants me to give the most information about Linda I can reasonably assume”.
There are variations in experiments concerning the Conjunction Fallacy, and whether my knee jerk critique holds water would require more familiarity with the literature than I have. However, this delightful article by Phil Maguire, Philippe Moser, Rebecca Maguire, and Mark T. Keane, gives a more than adequate account of why the Conjunction Fallacy is hugely overblown, if it exists at all.
In the middle of the 20th century, psychologists were obsessed with identifying ways in which “human rationality” was deficient, but their results have, in backward assessment, tended to say more about their experiments and themselves than about the human condition. I encourage a healthy skepticism anytime you hear someone bring up psychological fallacies.