Page 1 of 1

SRMR and NFI

Posted: Fri Oct 04, 2019 10:01 am
by aznur
I am a newbie in using PLS SEM. Seek help from the experts.

As attached is the results of my model fit. The results show:

Question 1

SRMR = 0.139 above than 0.08 (value suggested by Hu & Bentler, 1998)

Is there other cut off value for SRMR ?

Question 2
However, NFI = n/a ; while , Chi-square is reported as infinite .

What are the issues and possible remedies ?


Thank you
Aznur

Re: SRMR and NFI

Posted: Thu Oct 10, 2019 9:31 am
by jmbecker
Q2: Some of the model fit indices do not work (are not defined) for repeated-indicator models. There the same indicator appears more than once in the model and thereby causes perfect correlation in the indicator correlation matrix between the two repeated indicators.

Q1: The above might also be an issue here. If you want to assess/compare model fit you may want to go with two-stage higher-order constructs (assuming that the problem comes from repeated-indicator higher-order models).

Re: SRMR and NFI

Posted: Tue Jul 07, 2020 1:37 pm
by dennylarson
I have the same problem as u aznar
how do i get the value on the NFI and Chi