Testing the Reliability of Anchoring Susceptibility Scores

Whereas anchoring is a very robust and well-known effect that refers to the assimilation of numeric estimates toward previously considered numbers, the psychological mechanisms behind it have yet to be fully clarified. Research on theories on how susceptibility to anchoring is related to other personality parameters has not been able to provide sufficient empirical evidence of such relationships. A probable explanation is that anchoring scores lack reliability in most anchoring experiments. The present research examined whether reliability depends on the type of score used to capture anchoring susceptibility. In a classical anchoring experiment, participants (N = 78) were asked to estimate the true values of certain numbers (e.g., height of the Zugspitze mountain) after being confronted with either a high or a low anchor number. Four different anchoring scores that are commonly used to measure susceptibility to anchoring in anchoring research were computed for every person, as well as the scores’ reliabilities. The number and types of items were chosen to allow for reliable and valid measurement. Anchoring effects were present, but the reliabilities of all four scores were either very low or zero. These results reinforce the reliability problem that was also described by Röseler et al. (2022). There are probably no conditions under which anchoring susceptibility can be measured reliably, suggesting that there are no individual differences in susceptibility to anchoring. In further research, other person-independent factors that may influence anchoring strength should be investigated to develop theories that can explain the psychological mechanisms behind anchoring.