Timeline for Dropout in highly unbalanced longitudinal data (WGEE)
Current License: CC BY-SA 4.0
3 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| May 23 at 11:45 | comment | added | Frank Harrell | I would keep it simple and use a full likelihood method that only assumes missing at random (not missing completely at random like GEE, and that does not use an inefficient weighted analysis). Full likelihood methods include Markov processes, generalized least squares, and mixed effects models. | |
| Apr 23, 2021 at 13:47 | review | Late answers | |||
| Apr 23, 2021 at 14:00 | |||||
| Apr 23, 2021 at 13:31 | history | answered | teamug | CC BY-SA 4.0 |