Some great thoughts there Bernie, for a lovely sunny Friday here. :-)
> But I think you all get my drift...where in the process of designing a new "catalog" does user (patron) input come in? If the user's perspective doesn't come into play until the usability testing stage, that's probably too late.
I'll redirect your question to my library, since I think it serves as a similar example. We are a large public library, but with two large "research and reference" branches that think and act in many ways as academic libraries. It is a source of much institutional schizophrenia. Compounding the issue is that the proportion of "layperson" requests to "serious researcher" requests has not been well-quantified. Many suspect a 90-10 type of rule, but what generally happens then is that the advocates (usually, but not always, cataloguers) wind up forcing practice to revolve around the exception.
The *first* thing we need to do is stop lumping our users into a one-size-fits-all category of information-seeking behaviour. Use cases can be extremely varied, and we need to recognize that — for our institution anyway — there are *at least* two major personas we need to address (and quite likely an infinitude of more), and their needs are almost diametrically different. I use two not because of an affinity for bimodal models, but because one of the main concepts of access is the balance between precision and recall.
So, let's start there. Design one interface for high-recall, low-precision that will satisfy the (perhaps 90% of) users who want to search independently, without reference help, for that kind of use case. Design a more "expert" interface for low-recall, high-precision that will address the kind of detailed "serious research" the other users need. Such an advanced interface will almost certainly require expert training, but this is an opportunity both for professional reference librarians to flex their skills, as well as for education toward users who want to learn how to do it themselves.
Now, do this in the same system. Many will say that we already do this, but I think historically we've done a poor job of it (or, more specifically, the technology has done a poor job). Just looking at the recent increase in faceted search implementations points to one example of how the system can adapt to the user in a more organic way than we've had in the past. And maybe that's the by-and-for-librarians aspect that we mean; boolean search is powerful, but clunky. We can do better in 2010.
More importantly, we should design our metadata models and our cataloguing practices around this same notion — we already do it to some degree with copy-cataloguing (and eg. "levels of record completeness"); but looking at it as *intentionally* providing less description to meet the user need, not just as a way to avoid duplication of effort. One of the most frustrating phrases I see and hear on a daily basis is "good enough" — good enough for what? For whom? Practice needs to be tied to specific groups of behaviour, and one-size-fits-all doesn't cut it any more.
> I'm talking about an in-depth look at studies of information-seeking behavior in a higher education environment. And I'm not talking about studies of how library users use library catalogs (although that would be useful information). I'm talking about the whole information-seeking environment...not just the part that the library controls.
I *think* (but I don't know, not having done a lit. review) that there is a fair amount of research that could be pulled together to inform this. Maybe not comprehensive, and probably not as deep as we'd like for non-library information sources, but certainly this can't be an unexplored area. My curiosity is why isn't this research being distilled into practice?
> I've been involved in several projects to develop requirements for OPACs. Generally, when we'd come to the "end-user requirements" part of the process, we'd listen to public services librarians describe catalog functionality that they THOUGHT users needed.
Absolutely. As an anecdote, we're about to launch a (long overdue) new website interface for our entire library, including replacing the catalogue with a faceted search UI and some very un-librarian design. An intended-to-be-derogatory remark which one of my cataloguer colleagues made that I think resonates well is that "they want to make it look like a bookstore". It's probably one of the first things that we've designed for the 90%, and although public feedback has been overwhelmingly positive, the 10% — mostly librarians — have been extremely vocal about how they feel it erodes the sanctity and authority of the library's reference functions.
They may well have a point, but until we qualify and quantify the "serious researchers" next to our "general public" as a user group, someone is going to be left disappointed.
MJ
Received on Fri Apr 23 2010 - 13:31:05 EDT