Moscovici et al wanted to see if consistent minority could influence majority. His sample consisted of 172 females who were told they were taking part in a colour perception task, they were placed in groups of 6 with 2 of them confederates. In one condition the confederates were consistent with the incorrect answer and in another they were inconsistent with the answers. He found the consistent condition lead to 8.4% agreement, and the inconsistent condition lead to 1.25% agreement.