How many trials are needed in kinematic analysis of reach-to-grasp?-A study of the drinking task in persons with stroke and non-disabled controls.

The stroke results need caution because only 8 stroke participants were included, and test-retest reliability was analysed in only 5. Participants with stroke were selected for ability to hold an object like a remote control and attend 15 weeks of visits, so they may not represent people with more severe arm impairment. The results apply to the drinking task and to optoelectronic motion capture systems, not automatically to all clinical tests or all movement analysis devices. For non-disabled participants, movement units and inter-joint coordination had low between-person variation and ceiling or floor effects, which can make stability scores harder to interpret. Some trials were excluded because markers or movement patterns caused data gaps, so a practical session may need extra trials beyond the minimum.

Do not conclude that 2 to 3 trials are always enough for every movement variable. The returning phase needed 8 trials in the combined analysis, and peak velocity and time to peak velocity were less stable in the stroke test-retest subgroup.