CHI-WEB Archives

ACM SIGCHI WWW Human Factors (Open Discussion)


Options: Use Forum View

Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Ulrik Gade <[log in to unmask]>
Reply To:
Ulrik Gade <[log in to unmask]>
Wed, 27 Jun 2007 09:33:31 +0400
text/plain (130 lines)
Hi Anne,

My take:

First impressions are almost always relevant, as trained users will often be
desensitized to usability problems that it would still be valuable to fix.
And the basics are pretty much the same, in that a good study design exposes
a relevant mix of participants to a relevant mix of tasks in the UI. But I
take your point.

So how about simply "training" your test participants ahead of the actual
tests? Depending on your setup, you could simply start the session with some
key tasks that you first help them through then let them practice on their
own for a few minutes, then test with fresh data.

Or you could test each task twice; once without introduction (to get the
novice's perspective), then again after a thorough explanation and a bit of

Or you could train your participants ahead of the session, perhaps announce
a small prize for best test performance and then run the test at some
appropriate point when you can reasonably assume that their motivation and
skills are at their peak.

There are drawbacks of course, in that you may have less time for actual
testing of each task, and you'll have to work harder to have enough credible
test data available for both training and testing. But, hey, life is

Assuming the luxury of adequate numbers of participants I'd probably go for
a study design with a relevant mixture of experience levels. But you'll of
course need to set your own priorities.

Happy testing,

On 26/06/07, [log in to unmask] <[log in to unmask]> wrote:
> Hello all,
> We all know that we should do usability tests early and often in a user
> centered process.
> For "regular" web site, you take participants that are representative of
> your users, you make them do some tasks to validate ease of use.  Pretty
> straightforward.
> But for a web application?  Our users will be trained and will use the
> application as their main work tool.  How can we recreate that in an hour
> usability test?  We want to test the navigation and some main tasks.  But we
> think that the problems and comments we get in our current tests probably do
> not represent the "real life situation" of our end users.
> Of course, we get some valid comments on the application, but more, for
> example, on the position and visibility of the elements then on architecture
> and task support.
> Did any of you test the ease of use of a complex application with a
> standard usability test?  How can we detect major flaws before getting too
> far ahead in our design?
> Any comments or suggestions are more than welcome...
> Anne
> Anne Deputter
> ------------------------------------------------------------------------
> D.L.G.L. Ltée/Ltd <>
> Logiciels intégrés RH, Paie, Régime de retraite, Horairese
> HR, Payroll, Pension, Scheduling Integrated Softwares
> T : [450] 979-4646
> F : [450] 979-4650
> ------------------------------------------------------------------------
>     --------------------------------------------------------------
>            Tip of the Day: Postings must be in plain text
>      CHI-WEB: POSTINGS: mailto:[log in to unmask]
>               MODERATORS: mailto:[log in to unmask]
>     --------------------------------------------------------------

Ulrik H. Gade
c/o Marcura FZE
PO Box 54614
Office no. 204, West Wing 3
Dubai Airport Free Zone
Dubai, United Arab Emirates
Home: +971 (0)4 2888 147
Mobile: +971 (0)50 653 6181
Fax: +971 4 299 7923
E-mail: [log in to unmask]

Home address: Uptown Mirdif, Apt. 3106 (don't snail-mail me here, use the
c/o address)

The journey of a thousand miles starts with a single step

    Tip of the Day: Quote only what you need from earlier postings
     CHI-WEB: POSTINGS: mailto:[log in to unmask]
              MODERATORS: mailto:[log in to unmask]