Abstract
This study evaluated guilty knowledge detection using the Complex Trial Protocol (CTP) within the Guilty Knowledge Test (GKT) framework. Korean sentence stimuli were employed to examine the effects of countermeasures. The study involved three groups: a simple guilty (SG) group (n = 15), a countermeasure (CM) group (n = 15), and an innocent (IN) group (n = 15). Participants in the SG and CM groups engaged in a mock crime, while the IN group did not. Sentences were structured in a 'subject-object-verb' format, with crime-related information embedded in the object. The object element varied by stimulus condition (probe or irrelevant), whereas the verb ('steal') was constant across trials. Event-Related Potentials (ERP) analyses revealed late positive activity around 900 ms (P900) for probe verbs in both the SG and CM groups. However, source localization using standardized low-resolution electromagnetic tomography (sLORETA) indicated significant parahippocampal activation exclusively in the CM group. This suggests that the P900 in the CM group may reflect contextual recollection or memory modulation strategies employed to suppress recognition. Overall, the findings indicate that while CTP is effective in detecting guilty knowledge, sentence complexity and countermeasure strategies influence the expression and neural sources of late ERP components.