Standardized residuals in LPA PreviousNext
Mplus Discussion > Latent Variable Mixture Modeling >
Message/Author
 Kate Gliske posted on Saturday, November 18, 2017 - 11:11 am
Dear Dr. Muthen,

I need standardized residuals for my continuous latent profile indicators. The residuals I get with the output=residual command seem to be unstandardized. Is there a command to get the standardized figures with continuous indicators?

Best,
Kate
 Tihomir Asparouhov posted on Monday, November 20, 2017 - 10:31 am
You can use tech7 to get sample quantities for each class and model constraints to manually write particular quantities (along the lines of https://www.statmodel.com/download/StandardizedResiduals.pdf). As an alternative to tech7 quantities one can use the LPA model with unconstrained variance covariance in each class (that will give you SE).

Also using a command like this "define: standardize y1-y4;" will remove unwanted varying scales among the class indicators.
 Ann-Renee Blais posted on Sunday, January 28, 2018 - 6:29 pm
Hello,

I'm trying to run a LPA with 5 indicators (mean scores). The scores all vary between 1-5, except for one that ranges between 1-4. Should I standardize my indicators to remove the varying scales?

Assuming I have to keep the raw scores, what is the best way of graphing my data or interpreting the various profiles given the varying scales?

Thanks!

Ann-Renee
 Bengt O. Muthen posted on Monday, January 29, 2018 - 4:06 pm
Don't standardize. Use the Estimated mean option in the plot menu.
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: