Hi!
I have some trouble with the output when conducting a CFA via the SEMLj module in interactive mode. Since my data is ordinal (Likert scale responses) I am using the Roubust Weighted Least Squares estimation method (WLSMV). I want to interpret model fit using robust values. The output provides robust (and scaled) values in the "Fit indices" table for SRMR and RMSEA. But the corresponding columns are missing from the "User model versus baseline model" table, where CFI and TLI are shown.
I am using the current latest version of Jamovi that can be downloaded for Linux (labelled "28.2.0.0" in Jamovi, and "2.7.27" on Flathub) with the latest SEMLj module version ("1.2.8"). A colleague is running the analysis on the same data using Jamovi for macOS (version "2.7.38"), and the robust column appears in their output. Could there be a glitch in the Linux version, or do I need to do something to get the additional values?
Robust values for fit indices in the SEMLj module
Re: Robust values for fit indices in the SEMLj module
Hi everyone,
I am experiencing a similar issue with the SEMLj module when conducting Confirmatory Factor Analysis (CFA) using the MLR estimator.
Previously, I was able to obtain scaled and robust fit indices, including CFI and TLI, in the SEMLj output. However, these values are no longer displayed in the "User model versus baseline model" table.
I have tried several combinations of jamovi and SEMLj versions, but the problem persists:
Has anyone experienced a similar problem with MLR estimation on macOS?
Is there a known workaround or a particular SEMLj version that reliably displays scaled and robust CFI/TLI values when using MLR?
I would really appreciate any guidance from the developers or other users.
Thank you very much!
I am experiencing a similar issue with the SEMLj module when conducting Confirmatory Factor Analysis (CFA) using the MLR estimator.
Previously, I was able to obtain scaled and robust fit indices, including CFI and TLI, in the SEMLj output. However, these values are no longer displayed in the "User model versus baseline model" table.
I have tried several combinations of jamovi and SEMLj versions, but the problem persists:
- jamovi 28.x
- jamovi 2.7.38 (macOS)
- SEMLj 1.3.0
- SEMLj 1.2.8
- SEMLj 1.2.5
Has anyone experienced a similar problem with MLR estimation on macOS?
Is there a known workaround or a particular SEMLj version that reliably displays scaled and robust CFI/TLI values when using MLR?
I would really appreciate any guidance from the developers or other users.
Thank you very much!