The sample of participants that took part in this experiment consisted of 20. The experience was rather stressful for those who remained independent, and conformity was found to increase with the size of the unanimous majority ranged against the real subject, but only up to a majority of three against one, after which further increases produced no significant increases in conformity. The aim of this study is to investigate personality impression formation on a replication of Aschs (1946) 3 study. About 37 per cent of judgements on critical trials were conforming responses-the real subject going along with the incorrect unanimous majority. The task was deliberately made sufficiently easy to ensure that errors were virtually zero when people were tested alone, without any conformity pressure, but in the experiment there was only one real subject or participant in the group, usually sitting at the end of the row and therefore having to formulate judgements after hearing those of the other group members, and the rest of the group were accomplices of the experimenter, trained to give unanimous wrong answers on 12 critical trials out of the 18. They were shown 18 pairs of cards, and in each case were asked to say out loud, one at a time, which of the three lines on one card was the same length as the comparison line on the other. A group of people (usually seven to nine in Asch's original series) were seated around a table and told that they were to take part in an experiment on visual discrimination. Students completed attributed course evaluation forms three times: (a) after receiving fabricated warm or 'cold' information but before seeing the instructor (b) after a 30minute exposure to. A classic experiment on conformity introduced in 1951 by the Polish-born US psychologist Solomon E(lliott) Asch (1907–96) and subsequently used by numerous researchers. Solomon Asch's classic (1946) warm/cold research was replicated in this study with introductory psychology professors, using students' evaluations of teaching (SET) as dependent variables.