Confirmation bias in group information seeking and its implications for decision making in administration, business and politics

Discusses research on the confirmation bias and other selectivity effects in group information seeking, and explores the impact on group decision making. After a short discussion of factors that increase confirmation bias in individual decision making, information seeking processes in groups are described based on the authors' research with groups of two to seven subjects. Results revealed that a confirmation bias (1) occurs after preliminary group decisions, (2) is stronger in groups with homogeneous preferences (further increased by commitment and confidence), (3) is decreased by contrived dissent (i.e., devil's advocacy), though to a lesser degree than by genuine dissent (heterogeneous preferences), (4) does not, as expected, preclude sensitivity to the situational context, (5) increases under justification pressure, but only when the decision may prove objectively wrong, (6) is stronger in groups with a directive leader, and (7) is stronger when the decision is made by an elected group representative. Furthermore, implications of the above findings for ``real world'' group decision making are discussed and measures to increase reality testing in groups are derived.

0159937