Effect size for within subject longit... PreviousNext
Mplus Discussion > Growth Modeling of Longitudinal Data >
Message/Author
 Rebecca Greif posted on Tuesday, September 24, 2013 - 1:19 pm
I ran an open trial treatment study (i.e., no control group) examining within-subject change on 4 DVS at 3 timepoints: pre-intervention, post-intervention, and 3 month follow up.

I analyzed the data using linear growth curve modeling but I'm unsure how to compute effect size (d) from the output. Can someone please assist me.

Many thanks for your help.
 Linda K. Muthen posted on Wednesday, September 25, 2013 - 11:55 am
One way could be to take the difference between the mean at the last time point and the mean at the first time point divided by the standard deviation at the first time point. You can find these values in the RESIDUAL output. You may want to post this question on a general discussion forum like SEMNET for other suggestions.
 Rebecca Greif posted on Friday, September 27, 2013 - 10:44 am
Hi Dr. Muthen,
Thank you so much for your reply. I have tried to follow your instructions; however, I dont see standard deviation of timpeoint 1 in my RESIDUAL output. Am I missing something or do I need to calculate the SD? If so, please advise about how to do that.

Thank you again for your help.

Best wishes,
Rebecca
 Linda K. Muthen posted on Friday, September 27, 2013 - 10:54 am
You take the square root of the variance which is on the diagonal of the covariance matrix.
Back to top
Add Your Message Here
Post:
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Password:
Options: Enable HTML code in message
Automatically activate URLs in message
Action: