I have a problem with the calculation of sample size of my study.
Design: each participant will value 4 health states, including one 'baseline' state and 3 new states of interest. The sequence of the 3 new states will be randomly assigned to each participant. This is similar to cross-over study with 3 treatments.
Outcome: The value of each state will be bounded to 0 and 1 (continuous in-between).
The study question: are there any difference between value of the 3 new states?
The results will be examined using generalized linear mixed model with patients as random effect.
With Regards