The Effects on Anchoring of Increasing Quantities of Disconfirming Evidence

Abstract

Anchoring has been shown to influence judgements in a wide variety of contexts, often in a dysfunctional manner (particularly when anchors are deemed unreliable). Identifying methods for mitigating the effects of anchors is important. Our experimental study utilises three abstract settings and arbitrary anchor values. We find strong anchoring effects in initial judgements. Providing disconfirming evidence of moderate helpfulness does, however, reduce the anchoring effects. Specifically, providing one or two items of disconfirming information is shown to have significant incremental benefits in reducing or even eliminating anchoring effects. However, surprisingly, providing three items of disconfirming information, in our setting, did not further reduce anchoring, suggesting some diminishing effect of additional evidence. This is consistent with prior research suggesting that individuals adjust until they are 'close enough', and then stop considering additional information. Our results have implications for a wide variety of judgement contexts, and the results are encouraging in suggesting that a relatively small quantity of disconfirming evidence could be sufficient for overcoming anchoring.