Timeline for generalized linear model with log link using log transformed fixed/random effects?
Current License: CC BY-SA 4.0
2 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Sep 9, 2019 at 20:07 | comment | added | user250071 | TimeLog is the log transformed time measurement. Again, while I appreciate your general input regarding this model, my question only concerns the pros/cons of having a log transformed fixed/random effect when using a log-linked model and not the areas you specified. I am wondering if it will be akin to log transforming an already log transformed variable or something. If the above model is not satisfactory, then apply the same question to a model that you believe is. Do you have any thoughts on this? I cannot find too much information on it in general. | |
| Sep 8, 2019 at 7:21 | history | answered | Dimitris Rizopoulos | CC BY-SA 4.0 |