Library Web Site Usability Testing Questions Summary Part 1

[Blake Carver]

I am happy to report I got quite a few good responses to my call for 
usability testing questions earlier this week. I am currently sifting 
through the long list of compiled questions to make some sense of them, and 
make the list more readable, I'll send that along to the list when it's done.

Today I thought I'd share the interesting comments I received along with 
the questions themselves, I found these comments to be very useful. It 
seems we may be thinking about usability in some new and very useful ways.



*What I am struggling to do is to learn not just what our site does right 
or wrong, but how people approach it, how they think, and what they do when 
the site doesn't work for them (aka how they cope with failure). It's those 
who fail that are most important to me, and the site needs to have some 
kind of fall back for those who get lost, or even more important, those who 
do not know where to begin.  So how do we know what most people are coming 
to the site for, how do they approach the site, and how do we make the site 
fail good, i.e. when the site fails for them, how do we make it so they 
find a new way to approach it.

* The other day at the desk, while talking with a couple other librarians, 
one remarked how it would be much easier to design a site if all the users 
were librarians (or just talked/walked/thought like librarians), to which I 
replied, better yet, if our web site would be much better if it was 
designed for the users, if it acted like a good reference librarian and 
worked through problems in their language, if it knew/walked/thought like 
they did.

*I'm not big on testing (the artificiality and demand characteristics alone 
make it rather worthless, I watch people from the desk, and if I see 
several people hesitating or clicking back and forth and not getting what 
they want, that means it's time to change something ("don't make the user 
think").

*The hard part for me was devising questions that weren't "leading," yet 
avoiding questions which were essentially testing the information literacy 
of the student, rather than the usability of the web site (e.g. "find 
scholarly information on gender and identity").

*Here, One thing we wrestled  with was how 'functional' to make our 
tasks:  can we test the web  page without simultaneously testing the 
background knowledge of our students?  Some of our questions simply ask 
students to 'look for  information on <topic>', and were intended to mimic 
real-life uses of the  page.  Of course, the difficulty then comes in 
interpreting the results  and turning them into concrete suggestions for 
improvement.

*I read  the that WE think users are interested in often aren't what THEY 
are interested in.  Of course, if you check web hits by page, you've got an 
idea of what is used the most, but not necessarily why.




Blake Carver
Web Librarian
The Ohio State University Libraries
carver.50 at osu.edu
   see also: http://www.lisnews.com