Association-based Concealed Information Test: A Novel Reaction Time-Based Deception Detection Method

• Published in: Journal of applied research in memory and cognition 2017-09, Vol.6 (3), p.283-294
• Main Authors: Lukács, Gáspár, Gula, Bartosz, Szegedi-Hallgató, Emese, Csifcsák, Gábor
• Format: Article
• Language: English
• Subjects: Clinical medical disciplines: 750; Deception; Female; Human; Information; Klinisk medisinske fag: 750; Male; Medical disciplines: 700; Medisinske Fag: 700; Memory; Neurology: 752; Nevrologi: 752; Reaction Time; VDP
• ISSN: 2211-3681; 2211-369X; 2211-369X
• DOI: 10.1037/h0101811

In recent years, numerous studies were published on the reaction time (RT)-based Concealed Information Test (CIT). However, an important limitation of the CIT is the reliance on the recognition of the probe item, and therefore the limited applicability when an innocent person is aware of this item. In the present paper, we introduce an RT-based CIT that is based on item-category associations: the Association-based Concealed Information Test (A-CIT). Using the participants' given names as probe items and self-referring "inducer" items (e.g., "MINE" or "ME") that establish an association between ownership and responses choices, in Experiment 1 (within-subject design; n = 27), this method differentiated with high accuracy between guilty and innocent conditions. Experiment 2 (n = 25) replicated Experiment 1, except that the participants were informed of the probe item in the innocent condition-nonetheless, the accuracy rate remained high. Implications and future possibilities are discussed. General Audience Summary In certain scenarios, such as legal cases or counterterrorism, it is of crucial importance to correctly detect deception. One of the potential technological aids under development is the reaction time (RT)-based Concealed Information Test (CIT). The RT-based CIT has very low costs and it is easy to implement: it can be run on any regular personal computer, it takes little time (10-15 min), and its results can be analyzed practically instantaneously. In a CIT, a person is repeatedly presented several items (e.g., personal names), among which one is a probe item (e.g., the name of an accomplice in murder) that only a guilty person will recognize, and consequently his/her responses will generally be slower to this item than to the other items. Consequently, a major limitation of the CIT is that it cannot be applied when an innocent person can be aware of this item-which is the main reason for its very sparse actual application in real life. In the present paper, we introduce an RT-based CIT that is primarily based on associations (and not on recognition): the Association-based Concealed Information Test (A-CIT). In our study, for probe items among the other items, we used the participants' own given names in the guilty condition, and randomly selected names in the innocent condition (as simulation for guilt and innocence in a real life case). The A-CIT included additional "inducer" items: words referring to the participants own given name ("mine," "my