Responsiveness, minimal important difference, minimal relevant difference, and optimal number of patients for a study