Sensorimotor Learning and error processing in a new Mirror-Reversal Paradigm: Proof of Concept and Preliminary Study
Yvain Longaretti, Abel Rassat, Ana Maria Cebolla Alvarez, David Zarka
January 9, 2026 preprint DOI: 10.31234/osf.io/ag2kx_v1 (opens in new tab)
Study at a glance
AI-extracted from the abstract| Characteristics | Preliminary experimental study |
|---|---|
| Sample size | 20 |
| Population | Human participants performing mouse-based reaching movements |
| Measures | reaction time (RT), trigger-to-target time (TTargT), movement breaks |
| Key findings | Mirror-reversed reaching produced marked increases in reaction time, trigger-to-target time, and trajectory breaks compared with normal reaching, which the authors interpret as heightened cognitive demand and explicit strategy use in de novo motor learning. Performance improved partially across blocks but did not return to baseline. The authors conclude the Flow Task is a valid multimodal tool for probing flexible motor control and conscious error detection. |
Abstract
Adaptability and flexible error correction are essential to interact with dynamic environments. However, previously learned automatic motor patterns can hinder performance under altered task demands, producing a “habit lag.” Mirror-reversal tasks provide a model for studying de novo motor learning, requiring suppression of prior motor plans and the formation of new mappings. Additionally, research in error correction of continuous mirror-inverted paradigms with high-resolution neural and behavioral measures remain scarce. Here, we introduce the Flow Task, a modified target-jump mouse-based reaching paradigm with left-right inversion, designed to probe error detection, correction, and conscious awareness, ultimately designed for seasoned meditators. Importantly, we present the Flow Task as an open-access, fully documented platform that includes the complete analysis pipeline and example datasets. For this preliminary study, twenty participants performed reaching movements under normal and reversed conditions while mouse kinematics were recorded at 100 Hz. We developed an algorithm to detect trajectory breaks, providing temporal anchors for future EEG analyses and microphenomenology inquiry. Primary outcomes included reaction time (RT), trigger-to-target time (TTargT), and movement breaks. Results showed marked increases in RT, TTargT, and trajectory breaks during mirror reversal, reflecting the heightened cognitive demands and explicit strategy use characteristic of de novo learning, with partial improvement across blocks but no return to baseline performance. This study demonstrates the validity of the Flow Task as a multimodal tool for probing flexible motor control and conscious error detection. By combining continuous kinematics with neural and phenomenological measures, this paradigm enables high-resolution investigation of when errors are registered, corrected, and brought into awareness.