Single subject study test force with:
- 2 Devices simultaneasly (A and B)
- 5 differents positions (s1,s2,s3,s4 and s5).
- 2 differents rooms (Y and Z)
Repeat each position 6 times consecutly. Different rooms test in different days.
Interested in:
- With the same device, can we find significance differences between rooms?
- In the same room and in the same device, can we find significance differences between positions?
- For the six repeated measures for each position, which kind of reliability option can we report? Or maybe is repeatability?
From the point of view of the study, I understand that the variability of the repeated measurements for each position should be less than the difference between position 1 and 2 so that the change between 1 and 2 could be considered real.
My idea is to calculate, if possible, the standard error of the measurement for the pairs of positions (s1-s2...), and add it to the average of the 6 repeated measurements of each position, and check if ranges overlap. If there is no overlap, we could be saying that the change is real.
For example, we have an average for the repeated measurements of position 1 of 250, and an average for position s2 of 240. The standard error of the measurement for the two rows (s1 and s2) of 6 data (6 repeated measurements) de gives us 9. With this we would have that s1 could be between 241 and 259, and s2 between 249 and 231. Since the ranges overlap we could say that there is no real difference.
