Cognitive processes during deception about attitudes revisited: a replication study.

Event-related potential (ERP) studies about deception often apply recognition tasks. It remains questionable whether reported ERP patterns and cognitive processes can be generalized to other contexts. As the study by Johnson et al. (2008) fills this gap by investigating deception regarding attitudes, we tried to replicate it. Participants (N = 99) were instructed to lie or tell the truth about their attitudes. We obtained the same results as Johnson et al. (2008): lies were accompanied by enhanced medial frontal negativities (MFN) and suppressed late positive components (LPCs) indicating that lying relied on stronger cognitive control processes and response conflicts than being honest. The amplitudes of pre-response positivities (PRP) were reduced for lies implying that lies about attitudes were accompanied by strategic monitoring. MFN amplitudes increased and LPC amplitudes decreased for lies about positively valued items revealing that lying about positively valued items is cognitively more challenging than lying about negatively valued items. As a new finding, MFN, LPC and PRP components were neither moderated by Machiavellianism nor by changes in the attitude ratings. The results indicate that LPC, MFN and PRP components are reliable indicators of the cognitive processes used during deception and that it is worthwhile to investigate them in further deception contexts.