Accessibility, Usability and the New Face of NINES

NINES stands for a Networked Infrastructure for
Nineteenth-century Electronic Scholarship, a
scholarly organization in British and American nineteenth-
century studies devoted to forging links between
the material archive of the nineteenth century and the
digital research environment of the twenty-first. In practice,
these efforts have centered on building a portal that
enables faceted searches of digital scholarly resources
that we have peer-reviewed. This web portal also allows
users to collect object and build exhibits online.
When NINES first launched its search interface in 2006,
the institutional challenges of aggregating scholarly
projects online were only the beginning. The NINES
search and collect interface, powered by a Collecting
and Exhibiting tool called Collex, was a new thing, relatively
speaking. Users were (and still are) very familiar
with the idea of search engines: Yahoo and Google have
long played a central role in any internet experience. But
faceted browsers, or tools that allow the user to continually
refine the categories of their search in order to pinpoint
resources from large pools of data, were in their
infancy. The most efficient of these were just beginning
to be implemented in the commercial sector: Amazon.
com is perhaps the best example of this. But the NINES
team needed to adapt the power and flexibility of a faceted
browser for scholarly research (rather than online
shopping) and design it in a way that communicated the
organization’s reputability as a peer reviewer and a forward-
looking software developer for humanities scholars.
See the result below.
A left-hand sidebar doubled as a user tag cloud and a
list of collected resources, while the central portion of
the interface was dedicated to searching and browsing. Searches in a faceted browser can become quite elaborate,
so an area was needed to keep track of the constraints
added by the user. The resources aggregated
within NINES also needed ample screen real estate, as
did the various genre designations of the objects for a
different kind of browsing. Search results were listed below
this frame, and were often “below the fold,” requiring
scrolling. The NINES interface was a difficult thing to build, and,
overall, it was a successful endeavor. Scholarly interest
in NINES grew, and the number of digital objects leapt
from roughly 70,000 to a whopping 300,000. Yet we
found ourselves with very little feedback about the usability
of this application. Were the targeted users (scholars
and students of nineteenth-century studies) comfortable
with NINES? Were they making use of the many
features Collex provided? The development team decided
to schedule a formal usability study to get answers.
As with any study of this kind, many of those answers
were surprising. Five helpful participants in the NINES
Summer workshop at Miami University demonstrated
the strengths – and weaknesses – of the Collex interface. As can be seen in this screenshot demonstrating the
eye movement during one of the usability sessions, the
user’s focus is scattered at any given moment. The tag
cloud at the left was an attractive feature, but its presence
alongside search results put it in direct competition with
them. Users frequently jumped back and forth from the
tag cloud to the summary of the search results and the
constraints. But once they found objects, users did not
think of creating an account to collect those objects until
they were prompted to do so by the test, and even then,
creating a free account gave them pause. Although each
person commented favorably on the variety and amount
of information integrated into NINES, every single person
tested said that the interface was a mystery – this
after having spent a full 20 minutes using it.
After some consideration of the results, we decided that
the fundamental structure of Collex was sound: its features
were compelling and the searching capabilities attractive.
In a way, Collex was too powerful, to the extent
that it was intimidating. Everything about our interface,
from the Home page to the Collex engine, had been conceived
in scholarly and theoretical terms, and demanded
a similar dedication from the user. And while NINES has
always sought to appeal to the growing community dedicated
to the mission of excellence in digital scholarship,
we never intended to limit our users exclusively to its
membership. We wanted Collex to be a friendly environment:
a place where both scholars and their students
go to conduct research in the nineteenth-century studies
online. As we began to sketch out the Collex redesign, our primary
goal was to break up the Collex interface into its
component features, giving each one its own important
area of the site. In many ways, our task was much simpler
than the one that faced the original designers of Collex –
in the intervening years, the number of faceted browsers
had multiplied considerably leading to the emergence of
web conventions for their styling and organization. Our
‘market research’ was directed to what worked and what
didn’t among the giants of internet commerce: Amazon’s
browser was the most intuitive, but Ebay’s method of displaying
search constraints was more informative. And, it must be remembered, a scholar is not always looking
for that one perfect object, as is the online shopper, but
rather a collection of objects that fit ever more specific
requirements. Collex needed to allow users to tunnel into
our data quickly and efficiently, all the while reminding
them of the numerous other materials available.
The new NINES home page was designed to communicate
two things: first, that this is a scholarly organization,
and second, that searching is an integral part of the
site. Making use of the familiar Google search blank,
we hoped to eliminate our site’s high bounce rate (the
number of visitors who leave almost immediately after
arriving) and invite those with merely casual interest to
stick around and explore.
The tag cloud – one of the most forward-looking and
dynamic parts of the NINES interface – was given its
own page, with plenty of room for browsing the interests
of NINES users. ool tips and help text were added (“What are tags?”)
whenever the use of features was subtle or required specialized
knowledge of social media software. And finally,
a space was reserved for a blog, so that users could
appreciate NINES as an active and ever-growing institution,
rather than a static website or finding aid.
But as these refinements were implemented, we kept
running into a logical disconnect between the kinds
of operations one could do in NINES. Searching and
browsing are available to any and everyone who visits
the site, and are easily found in the main navigation tabs.
However, what makes NINES unique is that is also a
scholarly workspace: after setting up an account, you are
able to collect objects, tag them, and re-mix them into
a essay or “Exhibit” of your own. NINES is first and
foremost conceived as a community for scholars, a place
where one can simultaneously contribute to and benefit
from an ongoing discourse on nineteenth-century studies.
Unfortunately, even with the new design, this aspect
of our mission was still not clear.
Taking a cue from Facebook, Flickr and other social media
sites, the NINES team conceived the “My 9s” page,
each user’s private homepage within NINES.
On this page, all your authoring, editing and collecting
efforts are centralized: as a user you have control over
how much of your personal information is shared with
other users while your tags, saved searches and recently
collected items are easily accessed. Exhibits, which had
previously been a mysterious and intimidating option
are now offered as a logical outcome of all your work
within NINES – why not share some of that hard work
with your peers? By enlarging the (pre-existing) social
networking attributes of NINES, we were able to fully
demonstrate the power and utility of Collex, all the while
making the many operations possible within it appear
more manageable and enticing. Going forward, we hope to incorporate other social networking
protocols to encourage the growth of a community
of NINES scholars. The “My 9s” page is an initial
step in this direction, and we are also planning a NINES
Facebook application, which will allow us to leverage an
existing social network to form our own. Because of the
way NINES is structured, we are only as strong as the
community that we are able to build. Part of this process
involves peer-review and aggregation via RDF metadata;
our summer workshops for new projects are another
important activity in this regard. But our web interface
and software tools are central to the ways that NINES
will gather scholars together. Through a combination
of usability studies, development meetings, discussions
with scholars in the field, and pure trial-and-error, we
have come to recognize the importance of re-tooling, an
ongoing process that is constantly engaged with the social
and technical ecology of the web.