Structural Constraints Shape Force Estimation: Dissociating Internal Representation and Cross-Limb Mapping.

Abstract

Force estimation is commonly assumed to rely on a unitary internal representation, with differences across paradigms attributed to task demands or feedback availability. The present study tests this assumption by examining whether force estimation reflects the combined influence of multiple structural constraints. Specifically, we dissociate the contributions of internal representation and cross-limb mapping using a contrast-based framework. Participants performed three force estimation tasks-ipsilateral reproduction (IR), contralateral matching (CM), and contralateral reproduction (CR)-across three force levels (10%, 30%, and 50% of maximal voluntary isometric contraction). Error measures (constant error, absolute error, and variable error) were analyzed at both the task level and the structure level. At the task level, IR and CM produced comparable patterns of systematic bias but differed in error magnitude and variability, indicating that the two paradigms are not functionally equivalent. At the structure level, contrast-based analyses revealed dissociable effects of internal representation and cross-limb mapping. Both constraints introduced systematic positive bias and increased error magnitude, particularly at low force levels, while their influence diminished as force increased. In contrast, no statistically significant modulation of response variability was detected under the present experimental design, although the limited number of repetitions constrains the interpretation of this finding. These findings challenge unitary accounts of force estimation and support a framework in which behavior reflects the interaction of multiple constraints whose influence is dynamically weighted by sensory reliability.