Federated Searching and Scaffolding

CogSci Librarian's latest post on the future of reference is unlike some of the doomsday and defensive prophesies I've seen floating around the library world. I highly recommend giving it a read and contributing to the discussion though your own posts or in the comments.

In response to her point about searching, I'll be glad to see the day when federated searches across databases from a variety of vendors can work smoothly. We have Metalib (our edition is called "Search Tools") at the University of Michigan, and while I teach it to intro-level classes, I generally avoid it when getting into upper division and graduate level courses because of the metadata that's available and similar across databases.

Metalib does a fine job with what it's working with, but until there's some standardization in terms of what fields are available to search and searching capabilities from vendor to vendor, we're being reduced to lowest common denominator when performing a federated search. I can't perform complex boolean searches because some of the databases searched during a federate search don't support them. Thesauri are database-specific in many cases, so controlled vocabulary is out of the question. I'm sure we'll hit the same problem when we try to make a federated search engine that does our catalog, our finding aids, our subscription databases.

This, in my mind, is cutting off any ability for us to help patrons construct great searches. If keyword searching is going to be as good as it gets with federated searching, I do think that I'm fairly useless when showing students metalib -- why even ask a librarian when they're not going to be able to get better results than your keyword searching?

I guess that's why I try to focus on evaluation of results and the importance of citing things properly when I teach lower-division college courses. Generally, the assignments they receive in COMM 101 call for only 6 resources. They can find 6 resources using keyword searching in Metalib. So, why focus on controlled vocabularies and complex combinations of Boolean operators when I can take a holistic approach to writing research papers? Why not spend time with them talking about evaluating the things they find? Why not help them use RefWorks to cite their stuff, and give them idea about how to paraphrase and incorporate their findings into a coherent paper? If they can find 6 items through Metalib, they don't need more than that in terms of 'power searching' ... yet!

It's those upper-division courses when their topics are a bit more focused and they need to find articles that may not fit into a single keyword search. It's the grad students who are going to be producing lengthy tomes of research -- much longer than my 101 students' 5-page papers. I'll save my schpiel on using databases' native interfaces and getting the most out of strategic searching for then. What my 101 students need is something else.

So! In response to CogSci Librarian: no, I don't think that federated searching dumbs down the search process. I think it places limitations on the control you have when searching. Many educational practitioners (myself included) use scaffolding -- slowly letting students become more and more independent. I don't want to dump strategic searching on them until they've become comfortable incorporating sources into their papers. I'll let Metalib handle the searching. But as they grow, and as they progress in their studies, we do need to remove the scaffolding.

I saw a poster at ACRL, and if any reader out there knows who made it, or remembers the title, I'd love to hear about it. Or maybe it wasn't a poster. Shoot, now I can't remember where I heard this, but it impressed me: instead of just doing a 'library session' for an instructor, they will provide a list of topics they could cover with the amount of time it would take to cover each one. For example, I sent my faculty this message yesterday:

Why schedule a library session this Fall? While your students may have had a session in using the library before, there's so much more I can help them with. Take a look at the topics here, and pick one or two for me to cover in a library session. This way, if you're pretty sure you student have been introduced to Mirlyn, I don't have to teach it!

- Popular vs. Scholarly vs. Peer-Reviewed
- Introduction to Library Resources (Mirlyn, Search Tools)
- Managing, Citing, and Incorporating References (RefWorks, Style, and Avoiding Plagiarism through effective use of quotes and paraphrasing)
- Finding Primary Source Material (newspapers, news broadcasts, transcripts throughout time and from around the world)
- Fitting Wikipedia (and Google Scholar, and the WWW) into Research Papers

...or, I could custom design a session for your class's special info needs. Build a library session into your course's Fall schedule and let me know when and where to be.

I didn't incorporate the time limits into mine -- I figured opening the channel of communication between my faculty and me would be a good starting point for a discussion of how to incorporate information literacy standards into the class. We'll see how this goes. I shouldn't have sent it on a Saturday. Now it'll be buried in a bunch of weekend spam and junk mail when they check for it.

Anyway, I thought this approach was an elegant solution for a problem I've noticed: faculty members know they students have had a library session before, so they don't think I should come teach again, because they've already heard it. This way, they can get an idea of what I can teach, and they can customize their session for their students' needs, and skip the intro stuff if they feel their students don't need it.

Yes! I wish I could remember who gave me this idea! I owe them oodles of coffee.

Anyway. Long post. Thanks for reading. It's been very interesting becoming part of the blogosphere over the last few weeks. Thanks for the comments and making me feel welcomed.

Eric