R-square PreviousNext
Mplus Discussion > Multilevel Data/Complex Sample >
 Keith D. Wright posted on Tuesday, July 07, 2020 - 7:28 am
I know this will be a vague question, without much details related to the model. But, what is the range of R-square standardized estimate? My thought is it would still be 0 to 1, because you cannot explain more than 100% variance. So, is it plausible to see in a model a standardized estimate of R-square of .957, this seems unusually high for a social science model. -Thanks
 Bengt O. Muthen posted on Wednesday, July 08, 2020 - 5:26 pm
Any r-square is between 0 and 1. Yes, that value looks suspiciously high. This may be due to multicollinearity.
 Keith D. Wright posted on Friday, July 10, 2020 - 7:35 pm
Hi Bengt, thanks very much for the response as I figured. I am helping a student get ready to defend a dissertation on 7/14/2020 and she asked for my assistance at the very last minute, and I had no involvement with her analyses.
 Franzi Kößler posted on Friday, September 04, 2020 - 1:38 am
I was wondering whether it would be correct to use the R-square provided in the output as an effect size for a two-level mediation model with mediations at both levels? I was also wondering what kind of R-square it is? Is it the same one as for regular mediation (i.e., the completely standardized indirect effect as outlined by Preacher & Kelley, 2011)?

Thank you in advance!
 Bengt O. Muthen posted on Friday, September 04, 2020 - 3:48 pm
R-square is given for each equation on each level using the total variance at that level as the denominator.
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