Comparing Is Believing: A Selective Accessibility Model of Judgmental Anchoring

Judgmental anchoring constitutes an ubiquitous and robust phenomenon. Nevertheless, its underlying mechanisms remain somewhat mysterious. We discuss four accounts that attempt to explain anchoring effects: insufficient adjustment, conversational inferences and numeric priming seem to be insufficient to understand the phenomenon. As an alternative, we propose a Selective Accessibility model. Drawing on the notions of hypothesis-consistent testing and semantic priming, the model assumes that anchoring effects are mediated by the selectively increased accessibility of anchor-consistent knowledge. Tests of the model's basic assumptions are reported and its implications are discussed.