Hi Donald.
I believe that you pointed to several ways in which my description of the
experimental design was inadequate. The description below is better, I hope:
Subjects=1..21
Tasks=1..6
Sessions=1..3
Methods=1..3
Each subject solve all six tasks. Each subject used all three methods. Each
subject participate in three sections, where one method is used for each
section. Each section consists of two tasks. Sessions is primarily intended
to capture the temporal order in which subjects were exposed to, and learned
from, the tasks.
Sub. Session 1: Session 2:
Session 3:
1 Method 1: task 1 and 2 Method 2: task 3 and 4 Method 3: task 5
and 6
2 Method 3: task 3 and 4 Method 1: task 5 and 6 Method 2: task 1
and 2
3 Method 2: task 5 and 6 Method 3: task 1 and 2 Method 1: task 3
and 4
...
and so on, with tasks varied, so that 1 and 2 are not allways together, etc.
To compicate things, there is a difference between even and uneven numbered
tasks, so that one session always consist of one even and one uneven task.
Another complication is that uneven tasks consist of six sub-tasks, but we
should propably leave this at the moment.
So perhaps the above description is a bit more complete.
Kind regards,
Kasper
Donald F. Burrill wrote in message ...
>On Thu, 9 Mar 2000, Kasper Hornb�k wrote:
>
>> I have a question concerning repeated measures analysis. I am not sure
>> of whether a linear model with a factor that varies as repeated measures
>> are taken (e.g., order or session) is identical to a repeated measures
> ^^^^^
>You mention "order" here but you do not include it in your model below.
>Did you intend to?
>
>> analyses. I'll detail the question below.
>
>> I have a within-subject study in which subjects used three methods to
>> solve six different tasks. The experiment is run in three sessions, each
>> consisting of two tasks.
> Are the sessions always in the same order, with the same two
>tasks in each session?
>
>> Three of the tasks are very different from the other three tasks.
> Does a session always contain two dissimilar tasks, or do some
>sessions (perhaps only for some Ss?) contain two similar tasks?
>
>> For analysing this experiment, I plan to use a model like Y[ijkl]:= u+
>> subject[i]+ task[j]+ session[k]+ method[l]+ e[ijkl],
>> possibly adding interactions between task, method and session. Is this
>> a repeated measures analysis or equivalent to a repeated measures
>> analysis?
> You have not mentioned separate error terms for the
>several factors of interest. As I understand "repeated measures
>analysis" (and experimental designs corresponding thereto), each repeated
>measure would have an error mean square (sometimes called a denominator
>mean square) that is formally equivalent to the subject-by-measure
>interaction. (This would follow from defining "Subjects" to be a random
>factor; ordinarily one would not consider "Subjects" a fixed factor.
>Presumably Task, Method, and Session are all fixed factors.)
> I gather from your description that Task, Method, and Session
>are all repeated factors: that is, all Ss get to do all tasks, in all
>sessions, using all methods; although the formal design --
>
> Subjects by Tasks by Sessions by Methods
>
>-- appears to be incomplete in ways that you have not explicitly
>described. And it is not clear whether the Ss may perhaps be nested
>within an Order factor (if the tasks within each session varied among
>Ss, e.g.).
>
>> If not, how should I analyse these data using SAS's repeated measures
>> option?
> I am not familiar with SAS, and cannot address this question.
> -- DFB.
> ------------------------------------------------------------------------
> Donald F. Burrill [EMAIL PROTECTED]
> 348 Hyde Hall, Plymouth State College, [EMAIL PROTECTED]
> MSC #29, Plymouth, NH 03264 603-535-2597
> 184 Nashua Road, Bedford, NH 03110 603-471-7128
>
>
>
>
>===========================================================================
>This list is open to everyone. Occasionally, less thoughtful
>people send inappropriate messages. Please DO NOT COMPLAIN TO
>THE POSTMASTER about these messages because the postmaster has no
>way of controlling them, and excessive complaints will result in
>termination of the list.
>
>For information about this list, including information about the
>problem of inappropriate messages and information about how to
>unsubscribe, please see the web page at
>http://jse.stat.ncsu.edu/
>===========================================================================
===========================================================================
This list is open to everyone. Occasionally, less thoughtful
people send inappropriate messages. Please DO NOT COMPLAIN TO
THE POSTMASTER about these messages because the postmaster has no
way of controlling them, and excessive complaints will result in
termination of the list.
For information about this list, including information about the
problem of inappropriate messages and information about how to
unsubscribe, please see the web page at
http://jse.stat.ncsu.edu/
===========================================================================
