The fundamental attribution error

A driver cuts you off and you conclude he is a jerk, not that he is late for something. A colleague misses a deadline and you think 'unreliable', not 'three other deadlines that week'. When you miss one, you know exactly which circumstances caused it. The fundamental attribution error is the habit of explaining other people's actions by their dispositions (character, ability, attitudes) while underweighting the situation that produced the action. Ross gave it the name in 1977. The founding demonstration is Jones and Harris's 1967 study, in which readers inferred a writer's private opinion from an essay even when told that the writer had been assigned which side to argue.

The mechanism is perceptual. The person is in front of you; the situation is background, often invisible, and takes effort to reconstruct. A dispositional label is also a cheap, stable predictor, so it feels useful. Jones and Harris's numbers show the effect is a shortfall, not a blank: the assigned essays shifted ratings less than the freely chosen ones, so readers did discount for the situation, just not enough. In organisations this becomes blame. A post-incident review that ends with 'the engineer was careless' has stopped exactly where the attribution error stops, and the system that let one careless moment take down production is left in place for the next person.

Whether the error is fundamental is disputed. Joan Miller compared Americans and Hindu Indians at ages 8, 11 and 15 and as adults in 1984. At the youngest age the two groups explained behaviour alike; with age, Americans increasingly cited general dispositions and Indians increasingly cited context. The dispositional habit is learned, and strongest in individualist cultures. Daniel Gilbert and Patrick Malone argued in 1995 for the more modest name 'correspondence bias'. The bias is also sometimes right: people do differ, and a label can predict. The discipline is to ask, before you reach for a trait, what situation would have made a reasonable person do the same thing, and whether you have checked.