Human Time Awareness and Feedback-Driven Improvements of Time Reproduction

Abstract

Time perception is critical for cognitive and behavioral functions ranging from speech to motor control. Learning to time durations can cause errors; therefore, recognizing and correcting the temporal errors is essential for individual timing self- awareness, which humans innately possess. Conflicting reports over whether humans can discern the direction of the timing error (earliness/lateness) can be addressed by introducing feedback in behavioral tasks. To better comprehend the extent of human timing self-awareness and how feedback modulates the process of learning to time, three experiments were conducted utilizing the computerized visual temporal reproduction task. Participants viewed a blue square for a set amount of time selected from a mixed set of durations ranging from 1.5-6 seconds and then re-created the square’s on-screen time using a keypress and received adaptive non-directional feedback for their performance. Each trial could be repeated following feedback, allowing a “re-do” to learn from the successes or errors in the first trial. In the first experiment, I tested twogroups of participants on versions where non-directional feedback was provided after every response, or not provided at all. Temporal estimates were more accurate and precise with post-trial non-directional feedback, revealing a metacognitive ability and tendency to adjust temporal responses. To examine the neural underpinnings of these previous behavioral findings, a fMRI-EEG study was performed with the same paradigm. Blood oxygen level dependent (BOLD) activation in the supplementary motor area (SMA), an area highly implicated in time perception and sharpening of temporal estimates, was observed as the durations was encoded and reproduced. Notably, significant EEG-informed fMRI activity in the SMA showed that the contingent negative variation signal covaried with the BOLD signal during the encoding phase. Additional BOLD activations were witnessed in areas associated with the brain’s performance monitoring system and the default mode network along with more timing-related areas parietally, frontally, and subcortically. The third experiment profiled an environment of temporal uncertainty in which every trial was not always followed by a redo opportunity. Two groups of participants were tested in settings where the frequencies of single (initial only) and double (initial and redo) trials varied. Group members receiving a low frequency of redo trials (80% single, 20% double) exhibited lower absolute temporal error and were more precise in their temporal estimates than participants allotted to the high frequency group. This demonstrated that both groups learned the underlying trial structure of the settings, adapted, and adjusted their temporal responses accordingly. Holistically, these studies offered deeper insights into timing self-awareness, the mechanism for improving time estimates through a redo opportunity and varied feedback, the region-specific neural responses to temporal judgements, and how we learn to time in uncertain environments.