Cokie Anderson, Associate Professor, Oklahoma State University
Cokie Anderson presented a methodology that she had used to test the usability of web pages developed at OSU. She described the traditional process of usability testing — designing tests, recruiting properly representative respondents, etc. — and stated that this kind of thing was not realistic given the situation in which most librarians work, i.e. no funding, no time, inaccurate results.
As a response to this Cokie presented the “common sense” approach to user testing, based on accepted theories of good usability (in this case Steve Krug’s Don’t make me think).
Cokie’s web project that she wished to test was OSU Electronic Publishing Center’s online version of the Encyclopedia of Oklahoma History and Culture. Usability testing was one of the conditions of the grant; traditional testing was tried, but provided no interesting results, so the common-sense approach was tried:
- Select a few people, who could be trusted to tell the truth, provide them with an incentive (a small gift)
- Ask for their first impressions of the site
- Get these people to think out loud as they navigate the site
- Video record the test for later review
Typically, this kind of user testing results in:
- A single start page from which all information can be accessed
- The removal of meaningless terminology (i.e “Librarian” — cf. “Romanian”, introductions, instructions)
- Making things self-explanatory, and conventional
Cokie’s talk provided some timely reminders that we still do stuff that we knew we shouldn’t years ago, and points to some stuff we should be doing now. Good stuff.
Improving the usability of a university library website: user research, analysis, design and usability testing
Catherine Brys & Morag Greig, University of Glasgow
This presentation revolved around a project in improving usability and accessibility of Glasgow University Library’s website and OPAC. The basic redesign was intended to shift from a site focussed on “library structure” to a site designed around user behaviour.
A lot of work was put into finding out different user types, user tasks and analyzing current problems. The methodology used here was interviews with library staff, quick polls, user panels and observation testing (videoed).
The main problem areas were found to be:
- What users wanted to get hold of was buried deeply in the structure of the site
- Library search was not intuitive
- Page structure not intuitive
- “Librarian” terminology
- Unattractive presentation, poor overview
It was also fond that users
- Spend most of their time performing just 25 tasks
- 50% search for a resource
- 50% of these are looking for a book
Three alternative redesigns were presented with different layouts and colour schemes. Feedback was received regarding how much information should be presented and which colours users preferred. It was found that the majority of people preferred consistent metaphors — use of layout elements, text, etc. in a consistent fashion — and that people wanted more integration of the OPAC into the main webpage.
This work can easily be seen in the context of NTNU Library, where we suffer from some of the very same issues, and user testing of this kind is highly relevant. We do, however, have a rather different set of challenges because we are entirely dependent on third parties: ITEA (university IT services) for server space and our content management system, BIBSYS for our OPAC, UiO and our consortium for our federated search and eJournal system, and probably more that I can’t think of right now.
Information architecture for internal users
Anne Welsh, Drugscope
Anne Welsh gave a presentation on how to design a successful information architecture (i.e the structural design of an information environment such as a web page) for internal (intranet) web pages. While this might seem superfluous because good information architecture is good architecture whether it’s internal or external, Anne presented good arguments that this isn’t necessarily the case.
An internal web page differs from an external web page in at least one very important respect: the external page is visited by people who have to be convinced that this page contains important, useful information; the internal page is visited by people who accept the validity of the web page (or do not have any option but to accept its validity). It is perhaps because of this, internal users have a higher threshold for how much information is presented.
It was also pointed out that internal users use intranets primarily for two purposes: finding information and performing tasks, for example respectively, finding out the official translation of a person’s title and submitting a travel and sustenance claim form. Consequently, users do not linger long on the intranet — they are in and out.
Given these facts, good design for internal users can be described in the following ways:
- efficient searching
- simple taxonomies
- few errors in the system
- few routes through the information
On the other hand, information does not need to be spoon fed to users in the same way; for example, a description that is suited to students is in some ways insulting to the intelligence of a librarian (who has a very detailed knowledge of library procedures).
Thus, the information presented on internal pages does not need to describe everything in detail, but rather present the necessary information accurately and concisely. The design of internal web pages can be process oriented, and built according to existing process structures that a user will understand because of their expert knowledge.
So good design is design that is moulded by an understanding of the users; of course, good design principles such as consistency of terminology and metaphor still apply, but much of the “sales pitch” and “catch all” aspects of good extranet design falls away.
As a practical solution to this problem, Anne presented some internal web pages built using a wiki; these web pages came from universities and professional companies, and demonstrated amply that a wiki can be a powerful tool for internal staff.
The benefits of wikis are that the information can be adapted and improved as understandings of processes change. The wiki allows questions to be asked and answered in real time, and changes the way we view writing and publishing for peers.
A very interesting talk all round.