Levels of consensus and majority and minority influence

Robin Martin, Antonis Gardikiotis, Miles Hewstone

Research output: Contribution to journalArticlepeer-review

Abstract

Three experiments are reported which examine the effects of consensus information on majority and minority influence. In all experiments two levels of consensus difference were examined; large (82% versus 18%) and small (52% versus 48%). Experiment 1 showed that a majority source had more influence than a minority source, irrespective of consensus level. Experiment 2 examined the cause of this effect by presenting only the source label (‘majority’ versus ‘minority’), only the consensus information (percentages) or both. The superior influence of the majority was again found when either (a) both source label and consensus information were given (replicating Experiment 1) and (b) only consensus information was given, but not when (c) only the source label was given. The results showed majority influence was due to the consensus information indicating more than 50% of the population supported that position. Experiment 3 also manipulated message quality (strong versus weak arguments) to identify whether systematic processing had occurred. Message quality only had an impact with the minority of 18%. These studies show that consensus information has different effects for majority and minority influence. For majority influence, having over 50% support is sufficient to cause compliance while for a minority there are advantages to being numerically small, in terms of leading to detailed processing of its message.
Original languageEnglish
Pages (from-to)645-665
Number of pages21
JournalEuropean Journal of Social Psychology
Volume32
Issue number5
DOIs
Publication statusPublished - Sept 2002

Keywords

  • consensus information
  • majority influence
  • minority influence
  • consensus
  • influence

Fingerprint

Dive into the research topics of 'Levels of consensus and majority and minority influence'. Together they form a unique fingerprint.

Cite this