Direction of Change PreviousNext
Mplus Discussion > Structural Equation Modeling >
 learningmplus posted on Friday, February 17, 2012 - 2:41 am
I ran 2 models in which I am looking at the change between anxiety and depression symptoms in the context of a treatment study. I ran the model twice where the flow of change goes from anxiety and depression (effects from anxiety at post to depression at post; and anxiety at follow-up to depression at follow-up and then vice-versa). I get a negative sign from anxiety at post to depression at follow-up and from depression at post to anxiety at follow-up. Both lagged effects are negative in sign. I am confussed because the contemporeneous effects from anx to dep and dep to anx at post and follow-up are positive in sign and the lagged effects negative. Does the negative sign for the lagged effects makes sense?? I pasted part of the output with one of the models wihout covariates.
RCMAS 0.118 0.168
CDI 0.142 0.095
RCMASPO 0.401 0.000
CDIPO 0.047 0.688

CDI 0.304 0.001
RCMASPO 0.521 0.000
RCMAS -0.080 0.325

CDI 0.004 0.953
RCMAS -0.009 0.904
CDIPO 0.474 0.000
RCMASPO -0.276 0.028
RCMASFU 0.635 0.000

RCMAS 0.414 0.000
CDI 0.065 0.427
 learningmplus posted on Friday, February 17, 2012 - 9:40 am
I should say that I am not sure of the interpretation, not the actual syntax or running the data.

I don't know if the negative sign means that lower levels of anxiety at post (RCMASPO = -.27, p < .001)) predict higher or more changes in depression from post to follow-up (CDIFU ON CDIPO) or if It is actually the opposite where I would say that changes in anxiety from pre to post (RCMASPO ON RCMAS) work their way up to predict changes in depression from post to follow-up, but this will actually be less (more change in anxiety, less change in depression). The former makes sense, but not the latter interpretation, hence why I am not sure what to make of the negative sign.
 Bengt O. Muthen posted on Friday, February 17, 2012 - 8:17 pm
This general interpretation question is better suited for the SEMNET discussion list. Include information about the contemperaneous correlation between the two processes.
Back to top
Add Your Message Here
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Options: Enable HTML code in message
Automatically activate URLs in message