Feedback-driven event-related potentials in conditional discrimination: insights from a matching-to-sample study

This study examined differences in event-related potentials (ERP) associated with the presentation of programmed consequences during conditional discrimination training in a matching-to-sample (MTS) paradigm. Electroencephalography (EEG) data were continuously recorded from n = 11 participants using a 64-channel wet-electrode system at a sampling frequency of 1,024 Hz. Three-phase MTS training and testing were customized using PsychoPy and featured 12 arbitrarily related abstract stimuli explicitly designed for this study. EEG data processing and averaging were performed using ASA-Pro v. 4.10 using 0.5-40 Hz band-pass filtration and automatic artifact detection. Time-locked epochs for ERP analyses utilized a 1,000-ms window with a 200-ms pre-stimulus baseline; epochs were synchronized with electronic trigger codes associated with incorrect or correct programmed consequences following comparison stimulus selection. The difference between incorrect and correct feedback on mean ERP amplitude was significant, t(10) = -2.93, p = 0.015, d = -0.88, with a mean amplitude difference of 3.89 𝜇V (95% CI: [0.93, 6.86]). The effect on mean ERP latency was also significant, t(10) = -5.46, p = 0.0003, d = -1.65, with a mean latency difference of 109.2 ms (95% CI: [64.7, 153.7]). ERP amplitude differences were further associated with Phase III test scores, t(10) = -3.14, p = 0.005, d = -0.95, while their association with latency differences was not significant, t(10) = 1.46, p = 0.161, d = 0.44. Altogether, these findings underscore the sensitivity of ERP measures to feedback presentation during an MTS paradigm, lending new insight into the cortical neurodynamics during the establishment of conditional discrimination.