Timeline for Analysing a target based trial
Current License: CC BY-SA 3.0
5 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Nov 26, 2011 at 23:12 | vote | accept | argon1024 | ||
| Nov 16, 2011 at 23:42 | comment | added | Manoel Galdino | I never used any non-parametric model for real (i.e., beyond theoretical classes), so it's beyond my expertise. But my impression is that a non-parametric model is in general more complicated, and I guess so with multilevel models (probably even more complicated). Try with the parametric one and see if it fits the data. If not, then you may think to move on to more complicated things. But, again, I'm no expert on non-parametric stuff. | |
| Nov 16, 2011 at 21:37 | comment | added | argon1024 | PS What do you think about the suggestion of doing a "non parametric" multilevel model using ranks rather than values; I've read suggestions of this in various places, but nowhere I'd regard as authoritative? | |
| Nov 16, 2011 at 21:34 | comment | added | argon1024 | That's really helpful. This is essentially the solution I'd come up with here albeit in SPSS. But no-one in that thread was willing to confirm that I was going down the right path, so I went back to asking for more general advice, which has confirmed what I was thinking. Many thanks! | |
| Nov 16, 2011 at 1:48 | history | answered | Manoel Galdino | CC BY-SA 3.0 |