Zloteanu, Mircea (2023) Veracity judgements : reinterpreting the processes behind "deception detection". In: Annual Conference of the European Association of Psychology and Law (EAPL) 2023; 04 - 07 Jul 2023, Cluj-Napoca, Romania. (Unpublished)
Abstract
In its almost 100 years, deception research has produced conflicting findings regarding people’s ability to detect deception and has made little theoretical progress in understanding the judgement process behind veracity decisions. This stagnation stems from two sources: theory and methodology. First, the focus on behavioural “cues of deceit” has miscast the judgement process as a detection task when evidence for diagnostic cues is lacking, hindering theoretical development. Second, the use of poor statistical models has resulted in a focus on spurious and misleading findings, impeding empirical development. The proposed solution is to reframe the research in terms of veracity judgments – elements that impact and shift judgement. This view aligns with modern findings of few (if any) reliable and diagnostic cues and the near-chance level performance human judges ubiquitously display. This reframing requires modelling the underlying process as judgement shifts (decision threshold) alongside accuracy (sensitivity to cues), and incorporating the variability between senders and judges. The author demonstrates across several deception datasets the utility of Bayesian Mixed-Effects Signal Detection Theory (SDT) models. The results are presented using both a Bayesian Estimation and a Bayesian Hypothesis Testing framework, demonstrating how deception detection findings can be communicated to measure phenomena and test different techniques. To move the deception field forward there is a need to address the theoretical underpinnings of research and adopt more robust analysis techniques. The current results provide a conceptual example of veracity judgement investigations and aim to disentangle the conflicting findings of past research. A call for the use of more transparent, reproducible, and open practices is also made given the practical importance of such research.
Actions (Repository Editors)
 |
Item Control Page |