For those of you who don’t know OAIster, if you have any reason to search for digitized primary sources, you should check it out. It’s a union catalog of digital library holdings. It’s chief asset is wonderfully descriptive metadata. And like with other collections of collections, I recommend searching OAIster to find which digital collections contain the kinds of things you’re interested in, and then searching or browsing those collections individually.

For those of you who know OAIster, you know that it recently stopped being its own unique entity and started being an OCLC-hosted entity. It’s now available on the FirstSearch interface and the WorldCat.org interface. (Here’s more on the history of the catalog.)

Enter the oddness. My co-worker ran some identical searches on both interfaces and came up with startlingly different numbers of results for most of her searches. Confused, I contacted OAIster and have just heard back from them why this is so. Apparently, the “keyword” search in the FirstSearch interface searches through the Source, Subject, Title, and Notes indexes. The keyword search on the WorldCat.org interface searches all available fields and all indexes.

So now we know.

We’ve all heard that single search boxes are the only way to go when it comes to building search interfaces. We’ve probably also seen students who will bypass all relevant information or links on a page and zero in on whatever looks like a search box. But I never put these two pieces of knowledge together before. Not, that is, until just this morning as I was driving in to work. This morning I had a revelation:

Every page with a search box is a “single search box” page.

We may gripe about clutter. We may grouse about having not enough guidance surrounding our search boxes. It doesn’t matter. For people who are primed to search, they will only see the search box anyway. The other stuff may as well not be there. (For those of you getting hot under the collar like I would be if I were reading this right now? Hang on, I’ve got something for you in a minute.)

Here’s my Parable With Two Screenshots. We have several lists of electronic resources on our library’s website, each of which has a “Search” and “Browse” function at the top.

We’ve gently corrected students who started entering their topic keywords into the “search” box, but haven’t been able to get rid of the box entirely. “Oh those kids,” we thought. “Desperately seeking search boxes again.”

Then last week I had a professor call me in consternation that the library systems were telling her there was nothing on her topic. Turns out… she had entered her topic terms into that search box.

Ok, ok, so that one’s legitimately confusing. We’ve realized that while the existence of the box is out of our control, the wording next to it isn’t. Soon it’ll say something more like “find a database.” Still, this is only the first half of the parable, and probably not the most relevant half at that. It’s mostly relevant in that its proximity to the second half made everything come together in my head. And so, on to the second half…

For the past two weeks, I’ve also had students from a lit class coming to see me, all of whom want “something, anything from the last ten years written about [insert famous theme] and [insert famous piece of literature here].” Granted, searching for themes is hard. Even something standard like “performative identity” requires thinking up all kinds of synonyms (body, fashion, display, etc). But what struck me is that as soon as I set the date limiter on MLA International Bibliography, each student gasped in shock and surprise. This limiter is not hidden. It is three lines, or 1 inch, below the search boxes. And yet it had been totally invisible to my students as they focused all their energies on those tantalizing search boxes.

So now back to my revelation (and those of you who’ve been thinking “But we simply can’t do away with advanced search pages! Single search boxes aren’t always the way to go!!” can tune back in now). Here’s what I now think: We can feel free to have advanced search pages on any interface that we think functions better with all of those options laid out. It doesn’t matter. People who only want a single search box will only see that search box anyway. People who want the options will see and appreciate the options. Everybody happy.

Thinking about the things that I like about Google or about library databases in comparison with each other after my last post, I realized that library databases need crazy-easy URLs. I don’t click through 2 or 3 layers of a website to get to Google. I type “goo” into my address bar, which fills in the rest, which takes me to Google. If I could type “MLA” into the address bar and get to even something as complicated as “MLAbib.csa.com,” life would be easier. Sure beats my current option:
http://www.csa.com.ezproxy.carleton.edu/htbin/dbrng.cgi?username=carl&access=[gobbledygook]&db=mla-set-c&adv=1

If I could also set a cookie that would authenticate me from my own home computer, life would be even easier.

Still need to work on the seamless access to full text part of the equation, though.

Tipping point: reached.

Up until maybe the middle of last year, it was pretty easy not to worry too much about the problems of doing “real” library research on the free web. “The kids are doing it” was a phrase that simultaneously helped us to worry about the state of information literacy in this web-ified era and to dismiss the problem as one that “the kids” would outgrow, like braces or a lisp or chicken pox, as they became better versed in scholarly research practices.

Well, it’s not just the kids any more, folks. Enough journal publishers have opened up their indexing and abstracts to the free web that it’s now possible (especially in some fields) to actually do “real” library research on the web. And so people are doing just that. This year, our new faculty orientation session brought questions about Google-friendly access right to our door-step in a big way, and part of this rather disorganized thread on FriendFeed brought it up again.

And yes, ideally everyone could use one nice, big, easy search mechanism to do everything from the most broad to the most narrow topic and then get instant access to the full text of whatever they find. Too bad that’s impossible.

Google is more familiar and forgiving, it’s faster, and there’s a lot of good stuff in it (particularly if you’re searching for something that hasn’t had any controlled vocabulary assigned to it, yet). But currently, disciplinary databases do a better job of collocating like items based on something more robust than the author’s choice in vocabulary and PageRank. Currently, disciplinary databases do a better job of allowing scholars to leverage their disciplinary vocabulary and a better job of helping novices stumble upon key vocabulary terms. Currently, disciplinary databases are the only things that can offer relief to my students who say that there are just too many false hits in everything from Google to JSTOR (free text search may be what they’re used to, but they’re often relieved to leave it behind as soon as they’re shown controlled vocabulary).

But all that aside, neither option does the “access to full text” piece of the equation very well. Unless your library subscribes to the publisher versions of pretty much every eJournal out there (an expensive proposition) Google can’t actually help you get to whole swaths of full text, and even then you’d have to be on campus or logged in to your library’s proxy server or something. And even if researchers are in a disciplinary database, they’ll still often have to step outside of that database to get the full text, and while a link resolver is a wonderful thing, it’s still a long way from being a perfect solution to this problem. Either way this lands you at the A-Z list figuring out if we have access to the particular issue of the particular journal you want.

I wish it were easier. I wish access issues didn’t make researchers jump to the conclusion that we’re “hiding” stuff from Google, or that we’re being unnecessarily silo-ish, or that indexing is over-rated, or that you have to do “complex” searches in library databases. I also wish that we could bringing together disciplinary databases in ways that allow easy cross-searching without giving up the time-saving specificity of disciplinary focus and vocabulary.

Right now it feels like we’re balanced precariously on that tipping point with a precipice on each side.