Hi all,
Thanks for the replies (including off list). I have since resolved the
discrepant results. I believe it has to do with R's scoping rules - I had an
object called 'labs' and a variable in the dataset (DATA) called 'labs', and
apparently (to my surprise), when I called this:
lmer(Y~X +
On Aug 18, 2010, at 1:19 PM, Johan Jackson wrote:
Hi all,
Thanks for the replies (including off list). I have since resolved
the
discrepant results. I believe it has to do with R's scoping rules -
I had an
object called 'labs' and a variable in the dataset (DATA) called
'labs', and
No, apologies (good catch David!), I merely copied the script incorrectly.
It was
lmer(Y~X + (1|labs),data=DATA)
in my original script. So my question still stands: is it expected behavior
for lmer to access the object 'labs' rather than the object 'DATA$labs' when
using the data= argument?
JJ
On 2010-08-18 11:49, Johan Jackson wrote:
No, apologies (good catch David!), I merely copied the script incorrectly.
It was
lmer(Y~X + (1|labs),data=DATA)
in my original script. So my question still stands: is it expected behavior
for lmer to access the object 'labs' rather than the object
On Aug 18, 2010, at 6:45 PM, Peter Ehlers wrote:
On 2010-08-18 11:49, Johan Jackson wrote:
No, apologies (good catch David!), I merely copied the script
incorrectly.
It was
lmer(Y~X + (1|labs),data=DATA)
in my original script. So my question still stands: is it expected
behavior
for
On 2010-08-18 18:41, Johan Jackson wrote:
Hi all,
I figured out why this was happening. It is because my actual code was:
lmer(Y~X + (1|as.factor(labs)),data=DATA)
In this case, the as.factor function looks for object 'labs' not object
'DATA$labs.'
Scope is something you hear about don't
Hi all,
I figured out why this was happening. It is because my actual code was:
lmer(Y~X + (1|as.factor(labs)),data=DATA)
In this case, the as.factor function looks for object 'labs' not object
'DATA$labs.'
Scope is something you hear about don't worry about until it bites you on
your ass I
Hello,
Setup: I have data with ~10K observations. Observations come from 16
different laboratories (labs). I am interested in how a continuous factor,
X, affects my dependent variable, Y, but there are big differences in the
variance and mean across labs.
I run this model, which controls for
One difference is that the random effect in lmer is assumed --
implicitly constrained, as I understand it -- to
be a bell curve. The fixed effect model does not have that constraint.
How are the values of labs effects distributed in your lm model?
On Tue, Aug 17, 2010 at 8:50 PM, Johan Jackson
9 matches
Mail list logo