Timeline for How should I model interactions between explanatory variables when one of them may have quadratic and cubic terms?
Current License: CC BY-SA 3.0
5 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Dec 8, 2014 at 17:36 | vote | accept | Bajcz | ||
| Nov 21, 2014 at 21:25 | history | edited | Frank Harrell | CC BY-SA 3.0 |
Expanded to answer follow-up question
|
| Nov 21, 2014 at 18:07 | comment | added | Bajcz | Thanks for this answer. I've never used splines before, but I think I understand your example. I have a few follow-up questions, if that's ok? 1. When looking at the anova results from ols, as in your example, what is meant by "All interactions" beneath a factor? That is, all interactions with what? 2. Will a similar approach be permissible in a mixed-modeling approach? I think I'm stuck with needing random factors. Is your example compatible with, for example, lme4? 3. Will this work if some of the interacting treatments are categorical? For example, what if X2 were a 2-level factor? | |
| Nov 21, 2014 at 17:51 | history | edited | gung - Reinstate Monica | CC BY-SA 3.0 |
added r code syntax highlighting
|
| Nov 21, 2014 at 17:37 | history | answered | Frank Harrell | CC BY-SA 3.0 |