Why we're better at predicting other people's behaviour than our own

Psychologists have identified an important reason why our insight into our own psyches is so poor. Emily Balcetis and David Dunning found that when predicting our own behaviour, we fail to take the influence of the situation into account. By contrast, when predicting the behaviour of others, we correctly factor in the influence of the circumstances. This means that we're instinctually good social psychologists but at the same time we're poor self-psychologists.

Across three studies, Balcetis and Dunning asked students to predict how they or their peers would behave in various scenarios. This included whether or not they or others would help a researcher clear up a knocked-over box of jigsaw pieces; donate part of their participation fee to charity; or cheat on a self-marked quiz. The relevant situational factors were, respectively: being alone or in a group of two to three; being in a good or bad mood (induced via funny or boring videos); having anonymity. Whilst some of the students predicted how they and others would behave in these situations, other students were actually placed in these circumstances and their behaviour was recorded. The predictions were then compared against the reality.

When predicting the behaviour of others, the students were shrewd "lay psychologists" and took situational factors into account. For example, in reality, people were 27 per cent less likely to help clear up the jigsaw when in a group than when alone. When predicting other people's behaviour, the students anticipated this: they said their peers would be 22 per cent less likely to help when in a group. When predicting their own behaviour, however, they didn't think it would make any difference whether they were in a group or alone.

It was similar with the charity donations and the cheating. In reality, students provoked into a bad mood gave 23 per cent less money to charity. And students given the cloak of anonymity cheated more. The students in the predicting role anticipated these situational effects (although they underestimated them) when considering the behaviour of their peers, yet they imagined that their own behaviour would be immune. They thought they'd give just as much money whether in a good or bad mood, and be just as likely to cheat, or not, regardless of whether they had the benefit of anonymity.

Another trend across all the studies was for people to overestimate their own altruism (judged against the average of how people actually behaved), but to estimate other people's altruism more reliably. This is consonant with a mountain of past research showing that we tend to assess ourselves in an unrealistically favourable light.

"The good news," Balcetis and Dunning concluded, "is that people display some level of insight into the ability of situational variations to shape potential actions that their peers will choose. The bad news is that people fail to realise, or choose not to realise, that this knowledge should be applied to predictions of their own behaviour as well."
_________________________________

 
Balcetis, E., and Dunning, D. (2011). Considering the situation: Why people are better social psychologists than self-psychologists. Self and Identity, 1-15 DOI: 10.1080/15298868.2011.617886

See also: We're unable to read our own body language (earlier Digest post).
Strangers to ourselves (Psychologist magazine article).

Post written by Christian Jarrett for the BPS Research Digest.