mixed model ANCOVA
PS it might be worth considering (fixed) interactions of Race, Sex, Emotion with Score as well, if you have enough data to make it feasible.
On Sun, Jun 28, 2015 at 10:06 PM, Ben Bolker <bbolker at gmail.com> wrote:
On Sun, Jun 28, 2015 at 11:45 AM, Ian Danilevicz <iandanilevicz at gmail.com> wrote:
Hy, I have read a post about mixed models (the link is below) and I still in doubt. https://stat.ethz.ch/pipermail/r-help/2013-December/364091.html the problem was:
* * 1 subject factor (random, between subjects) called Subject
*>* * 3 categorical within subjects factors called Emotion, Sex, Race *> * * 1 continuous covariate (**WITHIN subjects**) called Score* *the suggested models were:*ModelRT <- lmer(logRT ~ Race*Sex*Emotion + (1 | Subject)) or using lme from the nlme package: ModelRT <- lme(logRT~Race*Sex*Emotion, random=~1|Subject) However, where is the Score variable?
Please try not to send HTML-formatted e-mail to the list ... You're right, I was too hasty in answering that question. I think it should have been ModelRT <- lmer(logRT ~ Race*Sex*Emotion + Score + (Score | Subject)) (I'm a little bit suspicious of the original description, which says that Sex and Race are "within-subjects" factors ...)