Standardizing vs. rescaling indicators PreviousNext
Mplus Discussion > Structural Equation Modeling >
 Cynthia Yuen posted on Friday, March 30, 2012 - 4:15 pm
When indicators of the same latent are scored on a different scale, what are the pros and cons of standardizing vs. re-scaling these indicators? If you know of any related references, that would great too!
 Linda K. Muthen posted on Saturday, March 31, 2012 - 4:00 pm
It is not a problem if the indicators are on different scales. We recommend keeping variances of continuous variables between one and ten for convergence ease. You can use the DEFINE command to rescale by dividing by a constant. We do not recommend standardizing variables. No references related to indicators being on different scales come to mind.
 Cynthia Yuen posted on Tuesday, April 03, 2012 - 2:54 pm
Thank you very much for your prompt response! It was very helpful.

Just to clarify, though, if I have three indicators and two are on a scale of 1-5 and one is on a scale of 0-3, can I have all those indicators load on the same latent without re-scaling the items in any way (i.e., it's fine if the indicators are on different metrics)?
 Linda K. Muthen posted on Tuesday, April 03, 2012 - 4:13 pm
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