Timeline for Discrete-time survival model output doesn't match Cox proportional-hazards model - problem with input of data?
Current License: CC BY-SA 4.0
10 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jul 6, 2023 at 13:08 | comment | added | EdM |
I must have confused the Anova() display of the model with Mating_time as the predictor (1 d.f.) with that of the first model with Round as a predictor. There were already 8 d.f. associated with Round in that first model, meaning that it must already have been encoded as a factor variable. I'll edit my answer accordingly, while keeping a warning about the difference between encoding such variables as numeric versus categorical.
|
|
| Jul 5, 2023 at 20:42 | history | edited | Blanca | CC BY-SA 4.0 |
added 2603 characters in body
|
| Jul 5, 2023 at 20:12 | vote | accept | Blanca | ||
| Jul 5, 2023 at 13:41 | answer | added | EdM | timeline score: 1 | |
| Jul 4, 2023 at 20:26 | comment | added | Blanca | First, Lukas you're right. I don't know how I didn't spot this. I guess I was tired. EdM, fixed this and the output now looks more like I expected. Added a last question in the edited version. Feel free to comment. | |
| Jul 4, 2023 at 20:24 | history | edited | Blanca | CC BY-SA 4.0 |
Added new script from the changes proposed by Lukas Lohse and EdM
|
| Jul 4, 2023 at 18:10 | comment | added | EdM | In addition to the comment by @LukasLohse, you have to remove all data lines for an individual after the time of the individual's event. In a situation with at most one event per individual, someone who had the event is no longer at risk at later times and thus must be removed from analysis. Coding those later times after an event as representing either events or non-events is not appropriate. After an event, the individual is no longer properly considered part of the study. | |
| Jul 4, 2023 at 15:24 | comment | added | Lukas Lohse |
mating_time is the response. You need to swap it and round in your model. round isn't even binary.
|
|
| Jul 4, 2023 at 14:40 | history | edited | Blanca |
edited tags
|
|
| Jul 3, 2023 at 20:35 | history | asked | Blanca | CC BY-SA 4.0 |