Study of between-subject and within-subject variability of electromyography data and its intrinsic determinants for clip fitting tasks.

Knowledge of motor variability is an essential step towards understanding a subject's working activity. This study aims to assess within-subject and between-subject variability and the intrinsic factors that drive variability during a clip fitting task in the laboratory. Muscular activity is recorded on six muscles of each upper limb. Four metrics are used: two characterizing muscular load and two characterizing intra-clip fitting dispersion. Factor analyses and muscle-by-muscle analyses are conducted. Independent variables linked to subject characteristics and to task performance are considered. Factor analysis reveals the combined activity of the different muscles. The short and highly constrained task shows both within-subject variability and between-subject variability. The latter predominates. Each of the independent variables induces a variability which affects the muscular load metrics, although only a few affect the within-clip fitting dispersion. The muscle-by-muscle and factorial score analyses results are consistent.