|
|
| Message/Author |
|
|
|
|
| I have some repeated cross-sectional survey data that covers 15 years. One way of analyzing these data is via age-period-cohort analyses. Such analyses are controversial because of the identification issue resulting from the fact that age=period-cohort. A solution is a cross-classified random effects model the treats the period and cohort effects as random effects. Let's say I have a list of covariates I wish to add at level one. Can I then conduct a an LCA analysis in Mplus using all of the variables (APC+covariates) while still treating the period and cohort effects as random? |
|
|
|
|
| To work with random effects you need at least say 25 units across which they vary. |
|
|
| Back to top |
|
|
