Informational Influence

What it is

A person who is unsure what is true will often treat the judgements of others as data. Another person's view of how far a light has moved, which route is shorter or which product is better is evidence, and a sensible person weighs it. Informational influence names this use of other people as a source of evidence. It differs from normative influence, which works through the approval or expectations of others and does not require that they know anything.

The distinction comes from Deutsch and Gerard's 1955 paper in the Journal of Abnormal and Social Psychology, titled A study of normative and informational social influences upon individual judgment. The paper is cited here for the distinction it names. This page does not report its design or figures.

Consumer behaviour writing has built on the idea. One textbook in the vault, Consumer Behaviour, describes three strengths of group influence: informational, in which a person takes facts from group members; normative, in which a person conforms to win approval or avoid disapproval; and a third, identification, in which the group's norms become part of the person's own self-concept. It credits the three-part scheme to later consumer researchers and not to Deutsch and Gerard, so the two-part version is the original and the three-part version is an extension.

Jonah Berger's Invisible Influence discusses imitation in similar terms. Berger describes copying others as a route by which they shape what we do, in part because others carry information and in part because exclusion has a cost. The first reason matches informational influence and the second matches the normative kind.

In effect

Muzafer Sherif's 1935 experiments are the clearest illustration of informational influence in action. Participants in a dark room judged how far a stationary light seemed to drift. The light did not move and nothing in the room indicated distance, so no participant had an answer to check against. Judging alone, each settled on a personal range and a typical value, and these differed widely. Judging together, participants converged on a shared value, and members who had been in a group kept a norm close to the group's when tested alone later. The other people were the only measuring instrument available, and they were used.

Later work shows how the source of the information matters. In a 2005 study of mental rotation inside a brain scanner, Gregory Berns and colleagues showed adults the answers of four apparent peers, who were actors, and in a separate session the answers of computer icons said to run a simple algorithm. Errors rose from about 14 per cent with no group answers to about 41 per cent when the human group was wrong and about 32 per cent when the computer was wrong. A wrong human majority moved participants more than a wrong machine did. The authors linked conforming to the group with activity in an occipital and parietal network and read this as a sign that group answers may alter perception itself.

The 2014 study by Daniel Haun and colleagues bears more on normative influence than on informational influence. Children switched more when the demonstrators were still present, which the authors read as evidence of social motivation, and in the same design the apes mostly kept what they had learned themselves.

What it does not say

The label does not say that conformity of this kind is a mistake. Treating others' judgements as evidence is a reasonable response to uncertainty, and the dark-room participants had no better evidence available.

It also does not separate cleanly from normative influence in real settings. Deutsch and Gerard drew the line because the two can be pulled apart in a study, for instance by changing who is present, but most everyday situations contain both, and the same outward behaviour can come from either.

The figures on this page describe particular experiments. The scanner study analysed 32 adults, and the child and ape study used 18 children, 12 chimpanzees and 12 orangutans in its first part. They show what happened under those conditions and do not give a rate at which people accept others' judgements in general.

The three-part scheme from consumer behaviour is a later extension and is described here as the textbook describes it, without a claim about how it was tested.

---