Dare I leap in where angels fear to tread?  Certainly D. Burrell has beat
this question something fierce.

May I take a simpler (and probably naive) approach?  Suppose I say that
the 8 conditions are categorically different - some aren't higher on any
scale, they are simply different.  Then I assert loudly that the 10
repetitions are performed all at the same time, with one 'set-up' of the
specific condition.  If I see no trends (or ignore the ones I see) in the
time sequence course of the 10 repetitions, then I can go ahead (I think)
and set up a One-Way Analysis of Variance.  IN table form, that would be
8 columns, with n(i) repetitions in each.

If there was a variation due to a particular setup, away from the overall
average, then we could detect that variation between each condition's
average, and the overall average.  We would say this variation was "due
to" the specifics of the condition.  In fact, some of that variation
could be due to the details of how we set up the condition.  Those
details could include minor variations due to how it was set up, and
'hidden' changes that we are unaware of.  So be it.  To detect these
sources of variation as distinct from the variations due only to the
changes you thought you made (! :), you will need a different design.

then there is the intriguing possibility that your 8 conditions are not
simply "8 brands of automobile."  Instead, you just might be taking on a
designed experiment, with three factors.  This could be interesting!

If you now have 8 conditions, all related but different, your 10 reps
could be seen as a nice way to reduce variation in the individual
measurements, but you will back it out later in the analysis with an
estimate.

Could you do say 2 reps, change to a different condition, then do 2 more,
and keep this up until you have 10 sets of measurements altogether?  Then
you could estimate variation due to each measurement, and variation due
to changing the specific donations of a 'condition.'

Jay

Pingu wrote:

> It has been many years since my pschology degree and i can now finally
> say that the last remnants of all stats knowledge i once had have now
> vanished. Recently i have started running experiments again and need
> to analyse my data - but i cant figure out which analysis to run. Any
> help would be greatly, greatly appreciated.
>
> Basically i have 8 conditions, and within each condition i measured
> the time it took me to perform a task, which i repeated 10 times. I
> now want to analyse the data to see whether any of the conditions lead
> to significantly faster response times than the others. I am sure i
> remember doing some analysis which measured all conditions against all
> others, but i just cant remember what it was. I have SPSS but still
> nothing is jolting my memory.
>
> Can anyone help?
>
> thanks,
>
> DN
> .
> .
> =================================================================
> Instructions for joining and leaving this list, remarks about the
> problem of INAPPROPRIATE MESSAGES, and archives are available at:
> .                  http://jse.stat.ncsu.edu/                    .
> =================================================================

--
Jay Warner
Principal Scientist
Warner Consulting, Inc.
4444 North Green Bay Road
Racine, WI 53404-1216
USA

Ph: (262) 634-9100
FAX: (262) 681-1133
email: [EMAIL PROTECTED]
web: http://www.a2q.com

The A2Q Method (tm) -- What do you want to improve today?




.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
.                  http://jse.stat.ncsu.edu/                    .
=================================================================

Reply via email to