Dear members of the smartpls community, Professor Ringle, Professor Becker,
I have two questions that i would appreciate any help elucidate
According to Applying Multigroup Analysis in PLS-SEM: A Step-by-Step Process "If a construct does not pass the third MICOM step, and there is a significant difference in the composites’ equality of mean values and variances across groups, then that construct should be removed from the analysis"
1- Does this mean it should be kept in the model but out of the MGA analysis results?
2- When checking for Compositional invariance, I noticed that Step 2 failed for one construct, what is the best course of action? excluding it from the model for the MGA comparison? not perform MGA and study each separately?
My sincerest thanks!
Nathan Brown
Urgent for paper! MICOM - Equal composite mean values and variances not established
-
- SmartPLS Developer
- Posts: 1284
- Joined: Tue Mar 28, 2006 11:09 am
- Real name and title: Dr. Jan-Michael Becker
Re: Urgent for paper! MICOM - Equal composite mean values and variances not established
The book chapter seems to be incorrect or at least imprecise in that statement that you cite.
If your constructs fulfill step 2 (compositional invariance) then you can compare groups based on their standardized path coefficients in a MGA. However, you cannot combine them into an overall analysis without accounting for the mean differences (e.g., by adding a control variable).
If they do not fulfill step 2 you should also not exclude them from your model (because that would change the estimates), but you simply cannot interpret group differences for effects where the construct is involved (as predictor or dependent construct).
If your constructs fulfill step 2 (compositional invariance) then you can compare groups based on their standardized path coefficients in a MGA. However, you cannot combine them into an overall analysis without accounting for the mean differences (e.g., by adding a control variable).
If they do not fulfill step 2 you should also not exclude them from your model (because that would change the estimates), but you simply cannot interpret group differences for effects where the construct is involved (as predictor or dependent construct).
Dr. Jan-Michael Becker, BI Norwegian Business School, SmartPLS Developer
Researchgate: https://www.researchgate.net/profile/Jan_Michael_Becker
GoogleScholar: http://scholar.google.de/citations?user ... AAAJ&hl=de
Researchgate: https://www.researchgate.net/profile/Jan_Michael_Becker
GoogleScholar: http://scholar.google.de/citations?user ... AAAJ&hl=de
-
- PLS Junior User
- Posts: 4
- Joined: Sun Dec 09, 2018 8:19 am
- Real name and title: Nathan Brown
- Location: Chicago Illinois, USA
Re: Urgent for paper! MICOM - Equal composite mean values and variances not established
Thank you Dr. Becker, this is exactly what i needed to know.