Asch's Variables

Variables Affecting Conformity: Asch's Research

Solomon Asch (1951, 1955) conducted a systematic programme of research to identify the social conditions that increase or decrease conformity. Using a deceptively simple visual task — judging which of three comparison lines matched a standard line — Asch demonstrated that a sizeable minority of people would conform to an obviously incorrect majority answer. By systematically varying features of the experimental situation, he identified three key variables: group size, unanimity and task difficulty.

The Basic Procedure

In Asch's baseline study, a naive participant was seated among a group of confederates (actors briefed to give pre-agreed answers). The group was shown a card with a standard line and a card with three comparison lines; the task was to identify which comparison line matched the standard. On critical trials, confederates unanimously gave an incorrect answer. Across 12 critical trials, approximately 75% of naive participants conformed at least once, with a mean conformity rate of around 32%. This was striking given that the correct answer was unambiguous to control participants tested alone (error rate under 1%).

Group Size

Asch varied the number of confederates from one to fifteen. With one confederate, conformity was negligible. With two confederates, conformity rose to around 13%. Three confederates produced a conformity rate of approximately 32–33%, and adding further confederates beyond this threshold produced little additional increase. The relationship is therefore not linear: conformity rises sharply up to around three to four opponents, then plateaus. This suggests that beyond a certain critical mass, the informational and normative value of additional group members is minimal. One interpretation is that participants suspect collusion when groups are very large, reducing the group's perceived independence.

Unanimity

Asch introduced a fellow dissenter — another person who gave a different answer from the majority — in several variations. The effect was dramatic: the presence of just one ally reduced conformity from approximately 32% to around 5.5%, even when the ally's answer was not the correct one. This finding has important implications: it suggests it is the breaking of unanimity — rather than the provision of a correct model — that provides the psychological support enabling independence. When the majority is unanimous, participants perceive stronger normative and informational pressure; a single dissenter undermines both simultaneously.

Task Difficulty

Asch made the discrimination task harder by making the comparison lines closer in length to the standard line. Under these conditions, conformity increased. This finding supports the role of informational social influence: when the correct answer is less obvious, individuals are less confident in their own judgement and more willing to accept the majority's view as valid information. Greater ambiguity increases the epistemic utility of looking to others, making ISI a more compelling process.

Evaluation

Asch's research was methodologically rigorous and provided internally consistent findings across a range of systematic variations. The results carry strong face validity — the basic conformity effect is robust and has been replicated widely. However, the research attracted several significant criticisms. The laboratory setting is artificial: participants judging lines in front of strangers bears little resemblance to the conformity pressures of everyday social life, limiting ecological validity. The American sample in the 1950s may also reflect particular cultural norms of conformity; cross-cultural replications (e.g. Perrin and Spencer, 1980, with UK engineering students) have found substantially lower rates, suggesting that historical and cultural context moderates the effect. Additionally, the procedure involves deception and mild psychological stress, raising ethical concerns about informed consent and the right to withdraw that were not fully addressed by the standards of the time.