Temporal error monitoring: Monitoring of internal clock or just motor noise?

Abstract

Understanding how humans monitor and evaluate temporal errors is crucial for uncovering the mechanisms of metacognitive processes, linking the fields of time perception and metacognition. In a typical paradigm, participants self-generate a time interval and subsequently can accurately evaluate its error. The implicit assumption in the field has been that participants monitor temporal representations. Even though temporal error monitoring has been replicated numerous times, it remains unclear what kind of information participants monitor when assessing the just-generated interval. Here, we assessed two scenarios in which participants could monitor sources of variability in temporal error monitoring: the internal representation of duration (Clock Hypothesis) or just motor signal (Motor Hypothesis). We assessed temporal error monitoring by inducing different levels of motor signal in motor timing, with the expectation that these levels of motor execution would influence temporal error monitoring outcomes. The motor signal was manipulated by instructing participants to either use button presses or joystick movements to produce time intervals, allowing us to evaluate and report how different levels of motor execution signal affect temporal error monitoring. According to the Clock Hypothesis, the additional motor signal should impair the accuracy of temporal error monitoring. Conversely, the Motor Hypothesis posits that additional induced signal should enhance the accuracy of temporal error monitoring. In line with the Clock Hypothesis, error monitoring performance was enhanced in a condition with a lower motor signal. These results show that humans evaluate their errors based on an informationally rich representation of internal duration, supporting metacognitive abilities in temporal error monitoring.

Public significance: Temporal error monitoring emerged from the fields of interval timing, decision-making, and metacognition, positing that humans evaluate the sign and magnitude of their temporal errors. Here, we critically test whether participants assess their timing representations as such and whether they are aware of the correctness of these evaluations.