Positions, positionsXrooms, rooms, roomsXdevices, positionXrooms, positionXroomXdevices. But really I´m interested in positionXroomXdevices.

Thinking in this, could be interesenting compare results with Anova post-hocs and differences explain by the range of SEM? How can relation both conceptually? SEM is based on reliability index and standard deviation.

Repeated measures factors: 1x5 (position) + 1x2 rooms. Between subjects (Devices 2). Why is not a repeated measures Anova in this case?

But from another, more practical perspective, I would like to analyze the behavior of the 2 devices in rooms Y and Z. How? Going back to the beginning, I would like to know if the change between positions is real, that is, if this change is larger than the changes that are observed for records made with the same position. This is why I had thought about the SEM and the MDC. However, I don't know if it could actually fit this approach.

On the one hand, I would like to report a value that characterizes the differences between the 6 test-retest measures for each of the 5 positions. Right now I have thought about the standard deviation, but I don't know if with this design I could opt for some other measure of reliability. Right now I have done a repeated measures anova in which I am only interested in the posthoc to compare the same position of the same method between different rooms.