| |
Abstract:
There is a growing body of evidence from functional
neuroimaging and computational modeling studies indicating that the
Anterior Cingulate Cortex (ACC) is activated in conditions which
elicit a high degree of response conflict. The goal of this
particular study is to account for the finding of Braver et al.
(submitted) that the ACC is activated on low-frequency stimulus
trials in two-alternative forced choice, go-nogo, and oddball
tasks. We hypothesized that repeated presentation of the
high-frequency stimulus generates a response bias for that
stimulus-response mapping. When the low-frequency stimulus is
presented, participants must overcome the tendency to make the
incorrect response before making the correct response. This leads
to response conflict (i.e. ACC activity). Simulations were
conducted by adapting the neural network model used by Botvinick et
al. (in press) to account for ACC activity on error trials in
similar tasks. We incorporated sequential response biases into the
model by adding a small amount of priming to the appropriate
responses on each trial. As predicted, the measure of conflict in
the model was higher on low-frequency trials across all three
tasks. The model also captures key aspects of behavioral response
time and accuracy data, and generates novel and testable
predictions about ACC activity and behavior. The results suggest
that the conflict monitoring hypothesis may account for several
phenomena associated with performance of sequential choice
discrimination tasks.
|
