|
|
Contributors to model differences |
|
Message/Author |
|
|
We are studying personality change from pre- to -post disaster in a group of subjects on whom we have a continuous measure of disaster exposure. We have performed invariance tests comparing the two time periods. There was little difference between the pre- and post models. Nevertheless, we want to test whether the degree to which participants were exposed accounts for the degree to which the pre-disaster model differs from the post-disaster data. Is there a way to model the effects of exposure of model invariance? |
|
|
Are you thinking that more exposure would give rise to more measurement non-invariance across the two time points? If, so there probably are ways to analyze this using the Constraint= option of the Variable command together with Model Constraint, but it takes an advanced Mplus user to carry this out. |
|
Back to top |
|
|