Hi, What if in the results of MICOM, one out of six constructs does not have compositional invariance (step2). however, for all constructs the difference in means and log variance is insignificant. Can we say full measurement invariance is established and proceed with MGA?