Full Measurement Invariance and then?
Posted: Mon Jun 25, 2018 2:07 pm
Dear All,
I ran the MICOM for MGA and fortunately, I got the result as full measurement invariance. It means that the different groups can be pooled and moderator analyses are permitted. Here is the question... Does the result mean that I can potentially choose two different ways for the further analyses 1) compare path coefficients across groups or 2) merge the data and set a dummy variable as the moderator in the model? Or the second option (merge the data) has to be chosen due to full measurement invariance. By given the research design as an experimental setting (no treatment vs. treatment), I tend to choose path coefficients comparison.
A second opinion is really appreciated. Thank you!
I ran the MICOM for MGA and fortunately, I got the result as full measurement invariance. It means that the different groups can be pooled and moderator analyses are permitted. Here is the question... Does the result mean that I can potentially choose two different ways for the further analyses 1) compare path coefficients across groups or 2) merge the data and set a dummy variable as the moderator in the model? Or the second option (merge the data) has to be chosen due to full measurement invariance. By given the research design as an experimental setting (no treatment vs. treatment), I tend to choose path coefficients comparison.
A second opinion is really appreciated. Thank you!