On 2013-07-15, at 4:00 AM, r-help-request at r-project.org wrote:
> On 7/14/2013 3:05 PM, Rolf Turner wrote:
>> On 15/07/13 08:57, Spencer Graves wrote:
>>
>> <SNIP>
>>>
>>> You may know that "mixed effects" is another term
for "repeated
>>> measurements".
>>
>> <SNIP>
>>
>> I must of course preface this comment with an "I am no
expert"
>> disclaimer, but I do not
>> believe that this assertion is correct. It would be more correct, I
>> believe, to say that
>> repeated measurement models form a particular sub-class of mixed
>> effects models.
>>
>> Those who *are* experts may well correct me if I am wrong about this.
>
>
> Thanks for the clarification, Rolf. You are absolutely correct.
> "Repeated measurements" refers to the sampling plan. "Mixed
effects"
> describes a form of mathematical modeling. There are balanced repeated
> measures data sets where people do not have to worry about the
> subtleties of mixed-effects modeling. An experiment with multiple
> litters with several mice in at least some of the litters would call for
> mixed-effects modeling. This would include "repeated measures"
on the
> litters but not the mice.
>
>
> However, "I am no expert" on repeated measures /
mixed-effects,
> either
And you could have repeated measures within matched blocks (e.g., litters as the
matched blocks, and some repeated measure---say trials---on the individual mice
in each litter). H. D. Kimmel promoted such designs in the 1960s in a slim
volume describing the analysis of two such designs.
--
Please avoid sending me Word or PowerPoint attachments.
See <http://www.gnu.org/philosophy/no-word-attachments.html>
-Dr. John R. Vokey