[R-sig-ME] question about linear mixed models

[JOSE A ALEMAN]

Dear list users,

I have perhaps what may seem like a very simple question, given that I'm
new to R. I have been trying to fit a linear mixed effects model to a TSCS
dataset where I want to include random intercepts for country and year. I
have noticed that I get slightly different results depending on what
package I use (lme4 and nmle). I think this may have to do with the syntax
for specifying the random effects formula with nmle. With lme4, I type
something like this:

mixed.model <- lmer (y ~ x1+x2+x3 + (1 | nation) + (1 | year), data=data)

and R returns the following output for the random effects:

Random effects:
 Groups   Name          Variance   Std.Dev.
 year          (Intercept)   0.00            0.00
 nation      (Intercept)   9.40            3.07
 Residual                      2.42             1.56

With nmle, however, the formula would have to look something like this:

mixed.effects <- lme (y ~ x1+x2+x3, data=data,
      random=~1|nation+1|year, method="REML")

The results, however, are different, and this leads me to suspect that I'm
not specifying the random effects formula correctly since R returns the
following formula:

Random effects:
 Formula: ~1 | nation + 1 | year
 Structure: General positive-definite, Log-Cholesky parametrization
                                           StdDev     Corr
(Intercept)                          0.9003       (Intr)
1 | nation + 1TRUE             0.9649       -0.918
Residual                             2.0908


Why is this the case? Or am I missing something important?

Thank you,
Jose A. Aleman, Ph.D.
http://faculty.fordham.edu/aleman