Transcultural Validation of the Physical Activity Scale for Individuals With Physical Disabilities (PASIPD) in French: The PASIPD-Fr

For the translation, permission to proceed was obtained, and translation was conducted in consultation with the developers. The investigators proceed to a first translation of the initial scale from English to French with two French native bilingual physicians that resulted in a unique french translation. An English native bilingual physicians proceeded to a back translation from French to English. This version have been sent to the developers who agreed that this back translated version was faithful to the original scale. For the validation, the investigators used the international recommendations for cross-cultural translation and adaptation of questionnaires. Population was individuals with neurological disease (stroke, multiple sclerosis, Parkinson disease and neuromuscular disorders) with significant in physical disability. Patients were invited in an ambulatory setting. Data collected for every participant included demographic and general variables (age, sex, level, cause and time since neurological disease, walking capacities and aid), scores of a physical performance tests (10 Meter Walking Test, 10 MWT) and the 4 self-administered questionnaires: the Dijon Physical Activity Scale (PAS) the Activities-specific Balance Confidence (ABC) Scale, the Medical Outcome Study Short Form 12 " (MOS SF-12) and the Hospital Anxiety and Depression (HAD) scale. Face validity was assessed using verbal feedbacks from physician and patients, and the mean time to fill-out the scale. Criterion validity was assessed with correlation and discrimination analysis between the scores for the PASIPD-Fr and the Dijon PAS. Construct validity was assessed using the correlation with scores of instruments that measure various aspects related to and physical activity (convergent validity) such as the self-reported questionaires ABC, SF-12 and HAD scales and also the 10 MWT. For consistency and reliability, the investigators calculated the Cronbach α coefficient and calculated the ICC between 2 completions at a 2-week interval. Sample calculation was based on an expected ICC \> 0.9, which required 50 participants to reject the hypothesis that the actual ICC is \< 0.6, which corresponds to the lower limit for a "good" reproducibility.