I have done Metric invariance, its partially invariate. but the problem is i dont know how to do scaler invariance and i didn't find any video or book for step by step method. can you guys help me with that. ??
the most important question is: Why do you want to test it? Scalar invariance is requisite for latent mean testing nothing else.
To test for scalar invariance, you need to incorporate the means of the indicators and model a mean structure (incl. item intercepts and latent means). How this is done technically depends on your software.
For details on the background, allow me some self-ad:
Steinmetz, H. (2010). Estimation and comparison of latent means across cultures. In P. Schmidt, J. Billiet, & E. Davidov (Eds.), (pp. 85-116). Taylor & Francis Group: Routledge.
I am currently using amos for my analysis. My model includes culture as moderator and product choice as moderators. In this regards, I need to check the model fit across the both groups. So, to test model fit i had to test the latent mean and intercept of items.
no, if you have two (or more) cultural groups and you expect differences in EFFECTS across both groups, all you need is metric invariance. Equal intercepts or latent means are not part of the game. If you hypothesize mean differences in both groups, the case is different. But this is no moderator effect but an effect of culture on the variables themselves.