Summary of: "Public Service Strategies of the California Digital Library" John Ober and Laine Farley, CDL Tuesday, December 15 8:30-10:00 a.m. in 202 South Hall (the Flex Lab).
A tape of this presentation is available on reserve at the Media Resources Center with a call number PSTF 4. Due to technical difficulties this tape only includes John Ober's presentation.
If you would like a copy of the handout version given out by John Ober, which includes his Power Point presentation (summarized below), please contact Corliss Lee at clee@library.
Introduction by Roy Tennant
John Ober's career has included teaching at the SLIS, travelling to Africa as an ALA Fellow, serving as network resources librarian at ISIS, working as acting Director of Systems at UCB, helping to establish a library at CSU Monterey Bay, and finally landing at CDL. Laine Farley has been at DLA and CDL for well over a decade, first managing Melvyl User Services and now managing the CDL System Services Division. Their public service credentials speak for themselves, but one to highlight is that they both care deeply about serving our users appropriately and well.
John Ober
Mr. Ober began by saying that he would present three public service models, and a notion of an evolution through these three models. He mentioned that he feels like he is coming to public service "through the back door" (first through systems, then through experience at Monterey Bay).
The presentation followed several slides, which were on the front side of a handout.
Slide 1:
Guiding questions
- Who are our users
- How can we consistently assist them
- What is the institutional context for personal assistance
Slide 2:
Intermediary Model
Matching user to info. Here's what we have.
- Interpretation/discovery of need is paramount
- Staffing desks
- Staged or tiered reference
ARL statistics were shown. The ARL webpage lets a person choose two variables and automatically graph results. Example shown was reference queries at UC campuses, showing a slight drop off in UCB reference queries between '95 and '97.
Slide 3:
Barrier-Help model
Enhance user self-sufficiency-Be aware of...
- user guides
- pedagogically aware systems (e.g., in Melvyl, "find more", "find fewer".)
- barrier removal (e.g., links to full text)
Examples of barriers to information discovery success were given:
Collection diversity
Information overload
Physical access
- Resources
- Technology
-
System hostility
Triple cognitive load
(example of a tutorial, user must not only learn what INSPEC is, but learn the tutorial and the technology used, ie. mouse, operating system.)
New technologies to address these barriers were also listed:
Search coach or avatar (cf: The Palace)
Could we do more to recognize online user confusion?
Possible indicators of confusion
- Page reloads
- Field mismatches
- Conceptual mismatches
- Etc...
-
Public service staff should be more deeply involved in identifying these barriers and the ways to address them.
Slide 4:
Architecture model (or lego model-building a sense of place)
Create info "portal" - Come hang out with us- create compelling infrastructure
- provide context
- (help) build tools
A quote from "The Mirage of Continuity" was shown and read, from a chapter from Richard Katz on managing information resources. The guiding principle that can be taken from the quote is that to continue to serve users well, given that they have all sorts of places to go, (including the Web), we have to build compelling, useful, connected-to-users , virtual, digital environments that include a sense of place.
A recent headline from Chronicle of Higher Ed announced: "High school students use web intelligently for research, study finds"
Example of My Yahoo was discussed. One may set up a personal profile, and certain information is automatically provided based on zip code and other information. (Related to portal idea)
The back side of the handout mapped components of the "Architecture" public service model to current CDL activities and strategies. For example, Mr. Ober is interested in the provision of "in case" and "on-demand" help through user guides, help screens, etc.
Laine Farley
Ms. Farley continued the discussion of current CDL public service strategy development in relation to the "Architecture Model of User Services".
She began by saying that one must be careful when providing help in a digital environment, because avatars and talking paperclips can make people feel uncomfortable. She is skeptical about the idea of the automatic recognition of confusion, feeling that a human is needed. Currently they get comments and help-line calls, and refer many requests to the campuses.
The trend toward customization seems to work better in some cases than in others. While Ms. Farley likes and uses Amazon.com, its algorithm for suggesting potential new books to the user does not work very well. On the other hand, Reel.com's recommendations are on target, but in this case, they are not based on an algorithm but on behind-the-scenes human input.
Ms. Farley announced that they are about to appoint a "Tools and Services" working group. Roy Tennant will be on it. They will be identifying principles, guidelines, and criteria, i.e., which kinds of tools and services should CDL be concentrating on?
- Tools that provide access?
- Tools to rank and filter search results?
- Tools to manipulate data?
- Tools to integrate resources?
- Linking to text?
- Database advisor?
- Tools that facilitate sharing and collaboration?
- Tools that customize?
She concluded that they need our help and comments re: what's important?
Slide 5:
Guiding Questions - redux
What service objectives - match, teach, create context,..?
How embed them in this environment?
Who are users and how are they known
How is quality of service know
Where are economies (of scale and scope)Q & A
Q-Apparent drop in reference questions at UCB: this should be viewed in light of the fact that reference desks were closed during this period, and that there was a tendency to redefine reference as "individual instruction".
A-L.Farley: Nation-wide, even taking this into account, there is a real drop.
J.Ober: ARL stats are criticized for being input measures rather than output measures. He showed another ARL graph showing "group presentations", which shows UCB increasing sharply, while other ARL libraries are steady. UCSD appears to be higher than UCB.
Q-What does CDL have planned re: user studies?
A-There are many possibilities, including user surveys, focus groups, ethnographic studies. Plans to do some focus groups on many campuses in late Feb, with help from P.Maughan. Plans to partner with SIMS.Q-There is a tension between providing a unique environment and an easy interface for users. Is CDL in discussion w/ other institutions to see that their approach is consistent & fits with emerging standards?
A-J.Ober: They focus on UC first, but are part of the DLF, which includes 22 members. The CDL Technologies Working Group is considering architecture and standards issues, and includes B.Hurley.Q-We used to have Stanford reps on HOS & other committees? Should we do this again? (Think of High Wire Press.)
A-Yes, this should be considered. Some working groups may even have industry reps.Q-Are you doing research on human-computer interface?
A-L.Farley: We have always.,BR> J.Ober: Don't want to say that they're engaged in research. Maybe applied research. They have no plans to contract a cognitive psychologist, for instance. They will talk to us instead.Q-A problem with personalized web pages is that users don't have one information need at a time.
A-L.Farley: A user can have as many profiles as they want.Q-Question of priority setting --what is more important, cutting edge or mundane tasks?
A-L.Farley: Need criteria to make these decisions. Using campus staff to help is good but takes more time. There is no question that they will support existing services, as of now.Q-What is CDL's direction and motivation re: scholarly publishing issues?
A-J.Ober: Contribution to scholarly communication is a high priority for CDL & Atkinson. Possible activities include a pre-print server, high quality new journals...Q-What is best done at campus level and what at CDL level?
A-J.Ober: CDL must identify some economies of scale and scope. Some work can be done both at CDL and campuses, and is not mutually exclusive. Some can be shared, such as user guides, tutorials. What about reference?
L.Farley: In the past, DLA was a service to the 9 campuses, while now it is moving toward a different, specialized library. Idea of the distributed reference model--CDL could build a tool, but couldn't possibly staff it.Q-Any distance education projects?
A-J.Ober: LSTA grant to host a database for 33 multi-type libraries. Involved in development of UC Merced, which is going to be more distributed than other campuses.Q-How do we ensure user understanding that it is the library doing all this?
(audience member): Maybe B.French should make license includes name of library in user interface of databases.Q-Tension of portal model is that we all want our (branch or library) page to be where people hang out.