Asch asked students to take part in a visual discrimination task. All but one of the ppts were confederates. The real purpose of the study was to see how the real ppts reacted to the behaviour of the confederates.
Procedure: 123 male US students were tested. Participants were seated round a table and asked to look at three lines of different lengths. They took turns to call out which line was the same length as a 'standard' line, with the real ppts always answering second to last. There was always a fairly obvious solution to this task, on 12 out of the 18 trials the confederates were told to give the same incorrect answer.
Findings: on the 12 critical tasks, the average conformity rate was 33%. One quarter of participants didn't conform at all, half conformed on six or more trials and one in twenty conformed on all the critical trials. When Asch interviewed his ppts afterwards, he discovered that the majority of them had continued to privately trust their own judgements, but changed their public behaviour to avoid disapproval from other members of the group (compliance).
Variables affecting conformity
Group size - very little conformity when the majority consisted of just one or two confederates. Under the pressure of a majority of three, the proportion of conforming responses went up to 30%. Further increases in the size of the majority didn't increase the level of conformity much more, so the size of the majority is important but only up to a point.
The unanimity of the majority - when the real ppt in Asch's study was given the support of either anothe real ppt or a confederate, conformity levels dropped significantly to 5.5%.
Difficulty of the task - the level of conformity increased when Asch made the lines much closer in length to each other. Lucas et al found that the influence of task difficulty is moderated by the self-effifacy (confidence in their own abilities) of the individual. When exposed to maths problems, high self-effifacy ppts remained more independent than low self-effifacy participants.
Evaluation of variables affecting conformity
Asch's research may be a 'child of its time' - Asch's findings took place in a particular period of US history when conformity was high. At this time, the US was in the grip of McCarthyism when people were scared to go against the majority. Perrin and Spencer repeated the study in the 1980s in the UK. In their initial study they obtained only one conforming response out of a total of 396 trials. In their subsequent study, they found levels of conformity that were similar to Asch's. This confirmed that conformity is more likely if the perceived costs of not conforming are high.
Independent behaviour rather than conformity - only one third of Asch's trials gave a conforming response. Asch believed that rather than showing humans to be overly conformist, his study demonstrated a tendency for ppts to stick to what they believed was the correct answer.
Cultural differences in conformity - Smith et al analysed the results of Asch type studies across a number of different cultures. The average conformity rate was 31.2%. It was 25% in Europe and the US (individualist cultures) and 37% in Africa, Asia and South America (collectivist cultures).