Timeline for How to correctly model repeated-measures random effects in a linear mixed effects model
Current License: CC BY-SA 4.0
7 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Mar 20, 2019 at 15:30 | answer | added | Isabella Ghement | timeline score: 7 | |
| Mar 20, 2019 at 13:28 | comment | added | Julian Potvin-Bernal | ReneBt, according to that page, this does appear to be the appropriate place. I've edited the title though, thanks. Roland, thanks for the input. Regarding REML, it's my understanding that if I want to compare models in an anova I need to use ML, and that ML produces slightly more accurate fixed effect estimates as well. | |
| Mar 20, 2019 at 13:27 | history | edited | Julian Potvin-Bernal | CC BY-SA 4.0 | added 89 characters in body; edited title |
| Mar 20, 2019 at 9:14 | comment | added | ReneBt | Welcome to CV. You question title is a coding question (Stack Overflow), but your final sentence does sound appropriate to CV, if you are wanting the rationalisation of nesting etc. See stats.stackexchange.com/help/on-topic to check if CV is appropriate, and if so please rephrase to make it clearer how it is statistical issues you want guidance on, as opposed to coding. | |
| Mar 20, 2019 at 7:15 | comment | added | Roland | General rule: A variable can't be at the same time be a fixed effect and the grouping variable of a random effect. Without knowing more, the first model could be reasonable. Don't know why you don't use REML for the fit. | |
| Mar 20, 2019 at 3:40 | review | First posts | |||
| Mar 20, 2019 at 9:14 | |||||
| Mar 20, 2019 at 3:35 | history | asked | Julian Potvin-Bernal | CC BY-SA 4.0 |