Timeline for Fisher Scoring Algorithm did not converge however model output looks fine in R
Current License: CC BY-SA 4.0
6 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| May 2 at 22:41 | vote | accept | Max | ||
| May 2 at 22:41 | comment | added | Max | Thank you both Adam and Noah, it was really a stupid mistake to write glm twice. Way to long discussion for such a silly mistake. Huge thanks to both of you, I was very confused but can now continoue writting my thesis. | |
| May 2 at 22:39 | comment | added | Max | Nevermind, now I see it | |
| May 2 at 22:37 | comment | added | Max | I honestly still don´t see the difference between the first and the second model, code, except for the standardized regressors. Could you explain more in detail what´s wrong in the Code? Thank you both for your input! I am aware of the events per variable problem as well as the issue with separation. My goal is to highlight these issues and my problem in the first model was that it did not converge (as stated in the warnings) but did not look odd to me regarding coefficients etc.. | |
| May 2 at 22:04 | comment | added | Noah | It's worth noting that the first model wasn't a logistic regression at all; due to a coding error, it was a linear regression. So it is irrelevant here. We never saw the logistic regression model without standardizing. | |
| May 2 at 21:22 | history | answered | AdamO | CC BY-SA 4.0 |