A statistical approach: you are comparing two conditions (old system
vs. new system) and have an idea about the direction of the effect
(the new system is more efficient, resulting in faster times). 

Can you make an estimate of the expected difference between the two
conditions? What is the smallest difference you would not want to
miss? It takes more measurements to detect a small difference. Look
up 'effect size' and 'statistical power'.

Are you planning to test both conditions on the same subjects or are
you planning to use two separate groups? For between-subjects
measurements there are other statistical tests than for
within-subject measurements. If everyone performs faster with the new
system (and you have been careful to control for learning effects and
other artefacts) you have a strong indication that the new system is
faster. But if you use two separate groups, there is a chance that
(by pure chance) the faster users end up in one group and the slower
users in the other. 

If you can clearly state your assumptions, determining the sample
size is pretty easy to do with basic statistics.

Marielle Winarto


. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Posted from the new ixda.org
http://www.ixda.org/discuss?post=31284


________________________________________________________________
Welcome to the Interaction Design Association (IxDA)!
To post to this list ....... [EMAIL PROTECTED]
Unsubscribe ................ http://www.ixda.org/unsubscribe
List Guidelines ............ http://www.ixda.org/guidelines
List Help .................. http://www.ixda.org/help

Reply via email to