Group-Induced Decision Polarization: Further Generalizations

Abstract

Three experiments on the group polarization phenomenon are reported. Experiments 1 and 2 found decision polarization on issues relevant to the South African socio-political context and among groups meeting repeatedly, over four sessions. Polarization in these more established groups was similar to that in typical single-session groups at the group decision stage, but established groups retained the degree of polarization to a greater extent. Nominal groups failed to polarize. Experiment 3 also found polarization with socially relevant items (attitudes towards women's equality) but failed to find differential effects owing to real and ‘symbolic encounters’. Overall, these results extend the generalizability of polarization, particularly to more established groups and with issues relevant to socio-political contexts. Additional findings concern group-size effects, methodological aspects, and recidivism of polarization effects.