Hello,
I am conducting a study examining the relationship between mental health profiles across three time points. The measure of psychopathology is the same at all three time points. The measure of well-being is the same at Time 1 and Time 2, but is different at Time 3 to reflect a more age-appropriate assessment of the construct. Is there an alternative to using latent transition analysis to examinate transitions (or the relationship) between profiles across time? From what I understand, LTA requires measurement invariance across time points, which isn't possible for Time 2 and Time 3 due to the different instrument. Is it appropriate to conduct a repeated measures latent profile analysis or is there a better method?
Any assistance would be greatly appreciated.
Have you examined (i.e., through separate latent class analyses for each time point) whether you get similar classes for the well-being items between the three time points despite the fact that the items are different at Time 3? In case some or all of the classes "look" similar or have a similar interpretation, I'd say it could still make sense to examine transitions from T2 to T3 even without establishing formal invariance of the classes. You do not need invariance to examine transitions. (However, it can make the interpretation easier.)
Thank you for your feedback, Christian. I will take this into consideration. I am in the process of conducting thr LPA. I just wanted to make sure I use an appropriate methodology.
- Badges
- Science method
- Similar topics
- Biological Science
- Ecology