Developing a cognitively plausible model of lie-truth judgements: An adaptive lie detector account

To date, no account of lie-truth judgement formation has been capable of explaining how core cognitive mechanisms such as memory encoding and retrieval are employed to reach such judgements. One theory, the Adaptive Lie Detector (ALIED: Street et al., 2016) is sufficiently well defined as to be implemented as a cognitively plausible computational model. Here we describe the first cognitively plausible account of lie-truth judgments and test it in three studies. Developed within the ACT-R cognitive theory (Anderson, 2007), the model provides a close fit to past data (Study 1), predicts novel learning data (Study 2), and generalises to more complex environments with multiple cues (Study 3). In so doing, the cognitive model provides strong support for the assumptions of an adaptive theory of lie-truth judgement, and substantiates a unique prediction of ALIED that lie and truth biases can be considered as functionally equivalent.