thanks.
well, basically. i ran this analysis once before, with just general linear
model, but nothing stuck after multiple comparisons.
the question is: do reading scores predict volume change over time (in
canonical "reading regions").
so i tried again, using linear mixed effects, adding subjects with only 1
time point, and with UID as the random effect--so it increased my "sample
size" so to speak. but i find nothing significant. no slopes are
significant, and if they are, they won't stick with MC.
but the degrees of freedom are odd, i think, for the data i have, and i
wondered if that was skewing results.
i don't know. feeling officially discouraged because it's odd that there
is NOTHING there...given what we know about brain maturation and how it
relates to emerging cognitive skill (i.e., reading). and maybe i'm just
not accepting that there is nothing there, and trying to blame on the
analysis i'm running.
On Sun, Apr 1, 2012 at 10:40 AM, bbolker [via R] <
ml-node+s789695n4523974...@n4.nabble.com> wrote:
> SHouston <suzannehouston <at> gmail.com> writes:
>
> > I am trying to run a linear mixed effect model on data. I have 17
> > longitudinal subjects and 36 single subjects, and this is the code I'm
> using
> > (below). So, INDEX1 is the column with brain volumns, and the
> predictors
> > are gort and age, by time ID (time they were seen).
> >
> > I believe my data is set up the right way, but when I run it, I get DF
> for
> > Intercept is 49, and DF for slope is 13? Why?
> >
> > lme.gort=lme(Volume ~ GORT_FLUENCY+AGE, random = ~ 1 | TIMEID, data =
> > subset(vol_data, INDEX1=='LH_FUSIFORM'), na.action=na.omit)
> >
> >
> > fit_vol_model1 <- function(df){
> > tryCatch(lme.gort <- lme(Volume ~ GORT_FLUENCY+AGE,
> > random = ~ 1 | UID,
> > data=df, na.action=na.omit), error=function(err) tag <<-1)
> > data.frame(Term = rownames(anova(lme.gort)), anova(lme.gort))
> > + }
> > models = list()
> > models$anova = ddply(vol_data, c("INDEX1"), fit_vol_model1)
> > summary(lme.gort)
>
> The rules according to which lme calculates degress of freedom
> are on p. 91 of Pinheiro and Bates 2000
> (also on Google Books, <http://tinyurl.com/ntygq3>). You should
> be able to work it out from there ... (It's not actually clear
> what your question was -- are you wondering why you have different
> df for different effects?)
>
> Follow-ups should probably go to r-sig-mixed-models <at> r-project.org
>
> ______________________________________________
> [hidden email] <http://user/SendEmail.jtp?type=node&node=4523974&i=0>mailing
> list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
> http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>
>
> ------------------------------
> If you reply to this email, your message will be added to the discussion
> below:
>
> http://r.789695.n4.nabble.com/Degrees-of-Freedom-for-lme-tp4522671p4523974.html
> To unsubscribe from Degrees of Freedom for lme., click
> here<http://r.789695.n4.nabble.com/template/NamlServlet.jtp?macro=unsubscribe_by_code&node=4522671&code=c3V6YW5uZWhvdXN0b25AZ21haWwuY29tfDQ1MjI2NzF8LTEwMDY2NTU1Mjc=>
> .
> NAML<http://r.789695.n4.nabble.com/template/NamlServlet.jtp?macro=macro_viewer&id=instant_html%21nabble%3Aemail.naml&base=nabble.naml.namespaces.BasicNamespace-nabble.view.web.template.NabbleNamespace-nabble.view.web.template.NodeNamespace&breadcrumbs=notify_subscribers%21nabble%3Aemail.naml-instant_emails%21nabble%3Aemail.naml-send_instant_email%21nabble%3Aemail.naml>
>
--
View this message in context:
http://r.789695.n4.nabble.com/Degrees-of-Freedom-for-lme-tp4522671p4524668.html
Sent from the R help mailing list archive at Nabble.com.
[[alternative HTML version deleted]]
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
