Temporal prediction errors modulate task-switching performance.

Area of Science:

• Cognitive Neuroscience
• Computational Neuroscience
• Neuroimaging

Background:

• Temporal prediction errors (PEs) modulate effective connectivity between the anterior cingulate cortex and the right anterior insular cortex (rAI).
• rAI activity is linked to efficient performance under uncertainty, suggesting temporal PEs might impair behavior-change capabilities.
• Previous research established neural effects but lacked behavioral validation of this hypothesis.

Purpose of the Study:

• To investigate the behavioral impact of temporal prediction errors on performance during task switching under uncertainty.
• To test the hypothesis that temporal PEs disrupt the ability to change behavior when unexpected demands arise.

Main Methods:

• Participants performed concurrent temporal prediction and a color comparison task with occasional change signals requiring behavior modification.
• Temporal prediction involved observing a moving ball striking a stationary ball with variable temporal gaps.
• Performance accuracy was analyzed in relation to temporal PEs and temporal delays.

Main Results:

• Behavioral accuracy significantly decreased as a function of both temporal prediction error magnitude and temporal delay.
• Findings indicate that increased temporal PEs and longer delays correlate with poorer performance in task switching.
• Results challenge existing explanations and support a predictive coding framework.

Conclusions:

• Temporal prediction errors disrupt behavior-change performance in perceptual uncertainty contexts.
• A predictive coding framework explains that failure to minimize perceptual PEs leads to task-switching deficits.
• The fronto-basal ganglia network, including the rAI, is implicated in processing temporal uncertainty and prediction errors.