Authors
Dagmar Scott Fraser, Massimiliano Di Luca, Jennifer Louise Cook
Published in
Experimental brain research. Volume 243. Issue 5. Pages 107. Apr 03, 2025. Epub Apr 03, 2025.
Abstract
The 'one-third power law', relating velocity to curvature is among the most established kinematic invariances in bodily movements. Despite being heralded amongst the 'kinematic laws of nature' (Flash 2021, p. 4), there is no consensus on its origin, common reporting practice, or vetted analytical protocol. Many legacy elements of analytical protocols in the literature are suboptimal, such as noise amplification from repeated differentiation, biases arising from filtering, log transformation distortion, and injudicious linear regression, all of which undermine power law calculations. Recent findings of power law divergences in clinical populations have highlighted the need for improved protocols. This article reviews prior power law calculation protocols, identifies suboptimal practices, before proposing candidate solutions grounded in the kinematics literature. We evaluate these candidates via two simple criteria: firstly, they must avoid spurious confirmation of the law, secondly, they must confirm the law when it is present. Ultimately, we synthesise candidate solutions into a vetted, modular protocol which we make freely available to the scientific community. The protocol's modularity accommodates future analytical advances and permits re-use in broader kinematic science applications. We propose that adoption of this protocol will eliminate artificial confirmation of the law and facilitate more sensitive quantification of recently noted power law divergences, which are associated with neurochemical disturbances arising from dopaminergic drugs, and in conditions such as Parkinson's and autism.
PMID:
40178611
Bibliographic data and abstract were imported from PubMed on 03 Apr 2025.
Read full publication at:
Please sign in
to see all details.
Advertisement
Stats
- Recommendations n/a n/a positive of 0 vote(s)
- Views 19
- Comments 0