Difference between MICOM and MGA results

[Capcap1668]

Dear all,

I am conducting MGA analysis by SmartPLS. I got the results of MICOM of the relationship between latent variable A and latent variable B are equal mean value and equal variance among two groups-> Full measurement Invariance of relationship between latent variable A and latent variable B among two groups.
However, when I conduct the MGA, the difference between path coefficients ( relationship between latent variable A and latent variable B) is significant.
Due to my understanding, if the relationship between latent variable A and latent variable B among two groups has Full measurement Invariance -> no difference between path coefficients ( relationship between latent variable A and latent variable B). Would it be true?

I am a little confused by this.

I would like to thank you for any explanation,

H.

[jmbecker]

As the name says, full measurement invariance in MICOM refers to the measurement model; not the structural model.
You have invariance in construct composition (basically weights), construct means and construct variances. This does not imply that you have equal path relations.