The role of attention control in visual mismatch negativity (vMMN) studies

Abstract

The detection of unattended visual changes is investigated by the visual mismatch negativity (vMMN) component of event-related potentials (ERPs). The vMMN is measured as the difference between the ERPs to infrequent (deviant) and frequent (standard) stimuli irrelevant to the ongoing task. In the present study, we used human faces expressing different emotions as deviants and standards. In such studies, participants perform various tasks, so their attention is diverted from the vMMN-related stimuli. If such tasks vary in their attentional demand, they might influence the outcome of vMMN studies. In this study, we compared four kinds of frequently used tasks: (1) a tracking task that demanded continuous performance, (2) a detection task where the target stimuli appeared at any time, (3) a detection task where target stimuli appeared only in the inter-stimulus intervals, and (4) a task where target stimuli were members of the stimulus sequence. This fourth task elicited robust vMMN, while in the other three tasks, deviant stimuli elicited moderate posterior negativity (vMMN). We concluded that the ongoing task had a marked influence on vMMN; thus, it is important to consider this effect in vMMN studies.

Stimulus expectations do not modulate visual event-related potentials in probabilistic cueing designs

“…We also have focused on evidence from predictive cueing designs in which there is a clear logic regarding how participants' expectations are manipulated. Effects attributed to expectation violations have been reported in visual oddball designs, although they are not always consistent across experiments using different stimulus categories (e.g., Kimura et al, 2009;Male et al, 2020;Feuerriegel et al, 2021b) or tasks (File & Czigler, 2019;Petro et al, 2023). In macaques, ES and surprise effects on electrophysiological measures have been observed in statistical learning designs that involve weeks of training and hundreds of image pairings (e.g., Meyer & Olson, 2011;Ramachandran et al, 2017;Kaposvari et al, 2018).…”