[R-sig-ME] lmer vs. nlme

[Ashley Cohen]

Hi All,

Unfortunately, as this data isn't published yet I can't give you the output
results.

However, the models are structured as follows:

model1 <- lmer(outcome ~ group + visit + group *visit + (1 | id), data =
data)
model2<- lme(outcome ~group + visit + group *visit ,
data=data,random=~1|id, na.action = na.exclude, correlation = corCompSymm())

The random effects components are identical in each:

Random effects:

Groups   Name        Variance Std.Dev.

id       (Intercept) 1.280    1.131

 Residual             1.271    1.127

Number of obs: 195, groups:  id, 59


Ashley


On Mon, Jan 18, 2016 at 3:25 PM, Thierry Onkelinx <thierry.onkelinx at inbo.be>
wrote:

> Dear Ashley,
>
> Please add some (reproducible) examples. Without that we can only guess
> which models you fitted and why the results are identical. Please do add
> the output of sessionInfo() as well.
>
> Best regards,
>
>
> ir. Thierry Onkelinx
> Instituut voor natuur- en bosonderzoek / Research Institute for Nature and
> Forest
> team Biometrie & Kwaliteitszorg / team Biometrics & Quality Assurance
> Kliniekstraat 25
> 1070 Anderlecht
> Belgium
>
> To call in the statistician after the experiment is done may be no more
> than asking him to perform a post-mortem examination: he may be able to say
> what the experiment died of. ~ Sir Ronald Aylmer Fisher
> The plural of anecdote is not data. ~ Roger Brinner
> The combination of some data and an aching desire for an answer does not
> ensure that a reasonable answer can be extracted from a given body of data.
> ~ John Tukey
>
> 2016-01-18 21:20 GMT+01:00 Ashley Cohen <ashleyhcohen at gmail.com>:
>
>> Hi Mixed Models group;
>>
>>
>>
>> I was wondering if you might be able to clarify some confusion I am having
>> with results from lmer.
>>
>>
>>
>> I am running a mixed model with fixed effects for treatment, time and an
>> interaction between treatment and time. A random effect has been added in
>> for subject (for repeated measures).
>>
>>
>>
>> I was under the impression, based on the response from Ben Bolker here,
>> that lmer would be estimating an unstructured correlation matrix:
>>
>>
>> http://stats.stackexchange.com/questions/86958/variance-covariance-structure-for-random-effects-in-glmer
>>
>>
>>
>> However, when I ran the model with lme, specifying a compound symmetric
>> correlation structure and random effect, I got the exact same results as
>> from lmer.
>>
>>
>>
>> I am not sure about what is going on and was wondering if you had any
>> thoughts on this. I also couldn’t find a way to pull out the correlation
>> matrix from lmer.
>>
>>
>>
>> Thank you,
>>
>>
>> Ashley
>>
>>         [[alternative HTML version deleted]]
>>
>> _______________________________________________
>> R-sig-mixed-models at r-project.org mailing list
>> https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models
>
>
>

	[[alternative HTML version deleted]]