Log in
RegisterFixed Effects Muddleties [General Statistics]
Hi ElMaestro,
IMHO you can't talk about inter-subject variance if using an "all effects as fixed" EMAphylistic model. In such a model there is only one error source, the residual error.
If one is interested in inter-subject variance (to what end ever) a mixed model evaluated via real mixed model software is indispensable.
Although it makes no or little difference in case of balanced datasets with no missings. I have played with Helmut's dataset:
BTW: @Helmut - No message / warning whatsoever during fit of the mixed model (subject(sequence) as random effect, all other fixed) with Proc Mixed.
❝ But then again why would anyone fit a 2,2,2-BE study with REML. It can be solved pseudo-exactly without an optimizer Al Gore Rhythm. And it involves all effects as fixed, even in the case of PROC GLM + "random".
IMHO you can't talk about inter-subject variance if using an "all effects as fixed" EMAphylistic model. In such a model there is only one error source, the residual error.
If one is interested in inter-subject variance (to what end ever) a mixed model evaluated via real mixed model software is indispensable.
Although it makes no or little difference in case of balanced datasets with no missings. I have played with Helmut's dataset:
CVinter CVintra
Proc GLM 9.9912% 49.1476%
Proc Mixed 9.9912% 49.1476%
GLM: s2inter = (MSsub(seq)-MSerror)/2; s2intra = MSerror
Mixed: variance terms directly read from Covariance parameters
CV = sqrt(exp(s2)-1)BTW: @Helmut - No message / warning whatsoever during fit of the mixed model (subject(sequence) as random effect, all other fixed) with Proc Mixed.
—
Regards,
Detlew
Regards,
Detlew
Complete thread:
- Intra subject variability vs Inter subject variability rana 2013-03-01 13:20 [General Statistics]
![[Mix]](https://static.bebac.at/img/mix.png "open in Mix view")
- CVinter < CVintra: happens sometimes… Helmut 2013-03-01 16:12
- CVinter < CVintra: happens sometimes… ElMaestro 2013-03-01 20:51
- CVinter < CVintra: happens sometimes… Helmut 2013-03-01 23:10
- CVinter < CVintra or negative variance component? d_labes 2013-03-02 15:57
- Sometimes both… Helmut 2013-03-02 18:46
- Must be the optimizer for Mixed Effects Muddles ElMaestro 2013-03-02 22:22
- Fixed Effects Muddletiesd_labes 2013-03-05 12:57
- Fixed Effects Muddleties ElMaestro 2013-03-05 13:06
- Fixed Effects Muddleties Helmut 2013-03-05 15:53
- Fixed Effects Muddletiesd_labes 2013-03-05 12:57
- CVinter < CVintra or negative variance component? d_labes 2013-03-02 15:57
- CVinter < CVintra: happens sometimes… Helmut 2013-03-01 23:10
- CVinter < CVintra: happens sometimes… ElMaestro 2013-03-01 20:51
- CVinter < CVintra: happens sometimes… Helmut 2013-03-01 16:12
Ing. Helmut Schütz