Feature-Dependent Sensitivity of Multichannel Surface Electromyography Amplitude Measurements to Analysis-Window Definition During the Five-Repetition Sit-to-Stand Test: Analysis Code and De-identified Derived Data

Analysis code and de-identified derived data supporting a comparison of two pre-specifiedanalysis-window definitions applied to the same five-repetition sit-to-stand surfaceelectromyography recordings from eight lower-limb muscles: the full acquisition span and asignal-derived analysis window. Includes trial-level and participant-level amplitude measurementsunder both definitions, window characteristics, trial-exclusion records, aggregate analytic-samplesummaries, the outputs of three post hoc analyses (a window-parameter sweep, a log-ratio model anda duration comparison), the analysis code, and a verification script. The verification scriptreproduces, from the shared data alone, the primary estimates, the mixed-model marginal means andthe post hoc log-ratio model; the window-parameter sweep and the duration comparison requirewithheld inputs (the raw recordings and the participant-level stopwatch summaries, respectively),and their code and aggregate outputs are included. Raw participant-level recordings are notincluded, because the informed consent and institutional review board approval do not authorizetheir public release; the shared materials therefore do not permit re-derivation of the analysiswindow or feature extraction from raw signals. Participants are identified by pseudonyms; there-identification crosswalk is not shared. Code is licensed MIT; data and documentation CC BY 4.0.