|
|
Data cohort & missing data |
|
Message/Author |
|
|
Is there an option available in Mplus to use the DATA COHORT command for an accelerated longitudinal design without losing all individuals who were not present at all waves (let's take the example of data with 3 waves and 4 cohorts)? After all, would it not be better to maximize the data that are available as usually done with FIML and include all individuals - even if they are only present at one or two waves? Thanks for your answer and your time! |
|
|
See the UG ex 6.18 multiple-group approach. |
|
Back to top |
|
|