ANCHORING EFFECTS AND THE APPLICABILITY O F ANCHOR-CONSISTENT INFORMATION - A SELECTIVE ACCESSIBILITY MODEL

Three experiments explored the mechanisms underlying anchoring effects - the assimilation of quantitative judgments toward a given standard of comparison. The implications of four theoretical accounts, which as cribe anchoring to insufficient adjustment, conversational inference, numeric priming, or selectively increased accessibility of anchor-cons istent information, were tested. Study 1 demonstrated that anchoring o ccurs for plausible as well as implausible anchors. This finding contr adicts the conversational account. Studies 2 and 3 showed that the act ivation of an anchor value is not sufficient to produce anchoring. Rat her, the occurence of anchoring effects depends on the comparison perf ormed with the anchor value. These results contradict the notion that anchoring is mediated by either insufficient adjustment or numeric pri ming. However, they are in line with the assumption that anchoring is based on the selectively increased accessibility of anchor-consistent information. The strength of the anchoring effect depends on the appli cability of this information.