Public Interfaces Committee
September 15, 2004
SML Lecture Hall
9:30-11 a.m.
Minutes: Julie
Out: Kalee, Debbie
1. Announcements (10 min)
Andy: The Science Libraries are poised to spend $1000 to redesign their web
site. The redesign will be mainly cosmetic, to make the site look more professional.
They'd like to end up with Dreamweaver templates.
Dale suggested a freelance designer from Utah.
Karen: We now parse the RSS news feed on our own server. We mulled the idea of
having individual library news feeds (science, social science, etc.), which would
require modifications to the News database.
Dale: French and German translations of the Welcome page are done. Arabic and
Russian translations are outstanding; Julie has contacted those curators but
now will follow up with Ann.
2. Orbis -- request to put back Call Number index on Simple Search screen
(10 min.)
PIC was asked by some cataloging staff to make the Call Number index display
on the Orbis Simple Search screen. (It was recently moved down in a scroll-bar
display, along with several other indexes.) The Call Number index is heavily
used by catalogers in their shelf-listing work. Following a long and lively discussion,
we decided not to display it. Considerations included:
--Whether it caused serious productivity problems for staff. We noted that
when an index is chosen for a session, it remains chosen -- thus staff do not
have to scroll and click it for every search. Earl noted that some staff to
whom he talked at first hadn't realized they had to scroll to see the Call Number
index.
--public vs. staff use of the OPAC. Because of Voyager limitations, staff
have to use the OPAC for shelf-listing. Nevertheless, the interface is primarily
for the public, and PIC's decisions about it are made for the public.
--original reasons for displaying only 5 indexes: our OPAC statistics show
that these are the top 5 patron-used indexes. User interface literature shows
that offering users fewer options increases their chances of picking the appropriate
option.
Julie will respond to the person who made the request explaining the decision.
The requestor had also noted that PIC did not ask for staff input on this decision.
PIC members felt that asking staff for input on every decision would seriously
slow down our work, and that in the past we have tended to ask for staff input
on major interface decisions. We agreed that since we are batching our Orbis
change requests, it's reasonable to invite staff to look at proposed changes
in torbis for a week-long comment period before PIC ratifies the changes.
3. User assessment (Katie) (45 min)
Katie introduced this topic by asking questions such as:
- When we make an interface change, do we have enough evidence to justify
that the change is solving a problem?
- What do we want to know about our site? Perhaps we could pick a major area
of the site to focus on. Over six months, we could concentrate on an area to
improve and design a way to gather evidence. One possible area is the various
search boxes on the site (Orbis, ejournals, Google site search).
- Should we start by talking to users about what they think is important?
- In thinking about what to test, should think about the "modularization"
of our site -- that is, readers approaching library content from various interfaces
(portal, MetaLib)?
Dale suggested a task-based approach to assessment. For example, we might take
a class syllabus and ask users to find the assigned readings (books, articles,
reserves, etc.) listed on the syllabus.
Katie thought that while that approach was useful, it might be too rigid and
limited.
We did think it could be combined with a "blue-sky" approach; we could
ask users "what would you wish for in looking for this information? what
would help you that isn't here?"
We debated whether to concentrate on student or faculty feedback; students
are much easier to get. It's likely that changes made to benefit students will
help faculty too. Nevertheless, we shouldn't just write off trying to get faculty
feedback. Karen and Katie are starting to gather faculty feedback for the portal;
those are open-ended conversations about what content people are looking for.
The results might be useful to PIC.
Where does MetaLib user testing fit in? PIC and the MetaLib implementation
group should work together to conduct it. There's no MetaLib implementation schedule
yet; Dale and Katie (who serve on the implementation group) thought it was likely
that there'd be an interface for PIC to consider by November 2004.
Rick shared some of the feedback from the Medical Library's recent Web site
user testing. Users suffered "search box" confusion because there are
so many different places on the site from which to launch a search; users are
happy to tell you what to improve.
Action: PIC members in public services departments/units will ask their
colleagues "what are the top 5 things we want users to be able to find and
effectively use on the library's Web site?" PIC members will synthesize
responses and share on pic-list.
Another action: We'll put a 3-5 question survey on the site, linked from News.
(A CCL survey done this way in late spring garnered hundreds of responses in
a very short time.) Questions will be along the lines of: your status; why are
you on this site; did you find what you wanted. Rick suggested a couple of other
questions: what would you add to this site; what would you remove. Katie will
draft the questions and send to pic-list. She will also talk to Danuta about
using "SurveyMonkey," a cheap online system that makes it easy to design
surveys and put the responses into a database.
3. Training -- tabled once again for lack of time!
© 2007 Yale University Library
This file last modified 10/24/06
© 2007 Yale University Library
This file last modified 10/24/06
Send comments to libweb@www.library.yale.edu
