Single-trial relationships between neural and pupillometric indices of
error-processing and behavior.
Abstract
The amplitude of the error-related negativity (ERN) is known to be
correlated with attention to task and general cognitive control
abilities. Yet, previous research has struggled to consistently link ERN
amplitude with behavioral accuracy or reaction time in the task from
which the ERN is being measured. This lack of relationship could be due
to many factors that are difficult to control for, so explorations of
other converging measures to understand error processing and subsequent
behavior adjustment are warranted. The current study examines how two
other physiological markers of error processing—the phasic pupillary
response (PDR) and the positivity following an error (Pe)— relate to
post-error behavior. Additionally, we also examine relationships between
the three physiological indices of error-processing. In the study, EEG
and pupillometry were simultaneously recorded while participants
completed 12 blocks (50 trials each) of an Erickson Flanker task. For
post-error accuracy, we found that on a single trial level, the
amplitude all three physiological error processing indices for error
trials predicted whether participants made a correct response on the
following trials. At the subject level, only the PDR predicted average
post-error accuracy. For post-error slowing, at the single trial level,
only the Pe predicted post-error slowing, whereas only the ERN predicted
post-error slowing at the subject level. This suggests that the PDR and
Pe might represent promising indicators of post-error behavior
adjustment, and also highlights the need to examine relationships at
multiple levels of analysis. We also found that both the ERN and Pe
correlated with PDR amplitude. Results are discussed.