Title: Human movement data - reliable enough for functional data analysis?
Authors: Lina Schelin - Umea University (Sweden) [presenting]
Alessia Pini - Università Cattolica del Sacro Cuore (Italy)
Abstract: In movement laboratories, advanced measurement systems are used to capture human motion, forces that cause motion, and muscle activity patterns during motion. A common feature of such systems is that they generate functional data. Human movement has an inherent natural variation and we cannot expect observed movement curve data to be identical when a task is repeated. Still, it is crucial that measurement tools are valid and reliable, i.e., that they consistently measure the quantity that they are supposed to measure. Functional data analysis methods are already being used for the analysis of human movement data. However, reliability studies are mainly performed on reduced data, such as specific events or features extracted from the functional data. A few works on reliability for curve data have been proposed in the literature, but with limitations and no clear recommendations. We present and compare methods identified in the literature for reliability assessment of functional data, both on simulated data and for an application to knee kinematic data.