I am going to use a longitudinal survey design (observational study) to examine the relationships between scale variables A, B, & C (i.e., A --> B --> C).
(1) What is the best interval between the two/three measurement points?
(2) Do I use measures for scale variables A, B, & C for each measurement point?
(3) What is the best data analysis approach to be applied?
I don't think there's just one answer to these questions. It has to depend on the clinical context. Maybe somebody's written a standard text book on the subject, but you'll still need to do some thinking to apply it to your own area of research. No short-cut!
The answer to question #1 can only be guided by content expertise. It should be based on the current available knowledge about the response time for each treatment (which I am assuming is your A, B, and C in this model, since you did not provide sufficient detail).
WRT question #2, it depends on what these variable are. It is not clear if ABC are interventions, observation periods, outcomes, explanatory variables. If you clarifying that for us, we can provide more details.
WRT question #3, it depends on how you answer #2, but based on your statement about longitudinal data, then I would argue you need a longitudinal model. It may also require a nested model (multilevel), but again, we need more information to provide helpful responses.
I don't think there's just one answer to these questions. It has to depend on the clinical context. Maybe somebody's written a standard text book on the subject, but you'll still need to do some thinking to apply it to your own area of research. No short-cut!
Not enough information to post an useful suggestion. Consider looking at:
http://uregina.ca/~gingrich/ch11a.pdf
- Badges
- Science method
- Similar topics
- Mathematical Sciences
- Statistics
- Applied Statistics
- Survey Design