Of Databases and Deep Structure

It’s getting to be library instruction season, so this is probably something we’re all hearing right about now:

Sure, I’m interested in bringing my class to the library to work on their research paper. Maybe you can show them a couple of databases that might be related to their topics?

Ladies and gentlemen, I submit to you that the answer should be:

Well, okay, it’s more complicated than that. Here’s why:

Several months ago, Kevin wrote a great post about the psychology of learning. One principle he mentions is “deep structure.” Deep structure is basically the essential components of something that can be used to draw analogies to something else, despite surface level differences. It’s a principle that comes from linguistics, but has been adopted to describe the underlying meanings in many other fields as well, including education (it’s particularly caught on in math). Recognizing the deep structure of something is a complex skill, one that often requires expertise in a specific area. In one classic example, students were asked to sort physics problems. Beginning physics students put the problems into groups based on commonalities in the problem statement (for example, they would put all the inclined problems together), while those with more experience grouped the problems based on the skills they need to solve them (Bedard and Chi, 1993).

There’s another famous example of this that comes from the Bedard and Chi study and which is recounted by Willingham. Students are presented with two problems, one about therapeutic rays targeting a diseased organ, and one about armies attacking a castle. At first, the two problems seem completely unrelated; even if students could solve one problem, they often couldn’t solve the other. The trick, of course, is that the two problems require the same solution; it’s just hard to see that because of the narrative gloss on the problem. Once students are trained to look for the deep structure beneath the irrelevant details, they are able to much more quickly solve the problem.

I would extend this to library databases: Beneath the EBSCO, ProQuest, or you-name-it vendor facade, most databases work in similar ways. That’s why librarians don’t collapse with anxiety every time we encounter a new database; we understand the deep structure of what’s going on, and can take what we’ve learned about databases before and apply them to a new situation. (Yes, there are some databases that require particular training and are sui generis; I definitely acknowledge that things like PubMed exist and require instruction.)

In first-year writing and similar courses, it seems to me that it is of value to attempt to expose some of the ways that databases work with students, in an attempt to get them able to transfer that knowledgein 75 minutes or less. Have we been set an impossible task? Perhaps.

Transfer, the literature tells us, is most likely to occur after continued exposure and practice with the knowledge in question. At the same time, in a given lesson, we can’t do too much and expose students to a laundry list of new material. So what if we taught deeper, and less? I think it might look like this:

Begin the class with something students are already familiar with—Kevin likes to use Google, and so do I. But you might also use another search engine, or even the library catalog, if your students have already encountered and used it before. Talk about what it’s indexing, and what it isn’t.

Then I would talk about how that search engine is different from a library database, and why that matters. I usually show a general database like Academic Search Premier, and we do side-by-side comparisons. We discuss facets and search string, and how that works. Then we’ll look at an item record and dissect what’s included, and how it turned up in our search. All this takes about 40 minutes.

Then, for the last 30 minutes of class, I direct students to our database list, and a) give them some suggestions of places to look based on their assignment or b) invite them to explore by discipline. We don’t talk about any of the databases explicitly other than the general one. I save a few minutes at the end of class for us to discuss the individual databases they used.

By talking about how a database works and explicitly stating that most databases work similarly, I try to strip away the surface differences and expose the deep structure. Then I ask students to apply that to a new context immediately. When I have taught multiple databases, I find that a) I repeat myself, b) people get bored and only use the database I talked about first or last, and c) no one explores any databases that we didn’t talk about. Exercises like this eliminate all three of those problems, as well as expose and demystify how databases work. Students do get exposure to multiple options at the end of class, but I don’t explicitly “teach” multiple databases.

How many databases feel right to you?



Filed under Bibliographic Instruction is Dead, Library Instruction, Posts by Dani Brecher

10 responses to “Of Databases and Deep Structure

  1. Kevin Michael Klipfel


    I learned a lot from this post; thanks for this!

    Question: if you were going to use Google and compare it to the facets in, e.g., Summon or Discovery or Academic Search – what would you do, exactly, w/r/t the side-by-side comparisons?

    I think that’s a great way to do it, but am wondering exactly how that might go down.


    • Dani Brecher

      Great question, Kevin! I actually meant that I compare the sorts of results that you get in Google and a library database, not the facets. Although that might be interesting, to talk about how the facets make drilling down to a desired result much easier than continuing to add keywords to a Google search…

      • Kevin Michael Klipfel

        Ahh … that’s interesting, too! And, yeah, that would be a really great way to talk about the facets in a more substantive, less pointing-and-clicking way.

  2. Kim Duckett

    Hi Dani,

    Thanks for this thoughtful post. Although I hadn’t been using the term “deep structure,” I strive to make these kinds of connections with students too. I think of it as trying to reshape someone’s mental model by providing greater context for what a tool is, why a tool is the way it is…or for that matter, why you use a library website to access a database or an online journal articles at all. For me, this context is more important than teaching an interface and it’s the heart of information literacy. Interfaces will change, but deeper structures change slower (I would say “don’t change,” but I’m wondering if tools like Google Scholar or discovery systems show ways that it can change…just a thought haven’t followed through to the end!).

    It seems to me that with the concept of “what is a database?” there are different angles you can get out to expose the deep structure. Kevin’s question to you seems to be about what I call anatomy, which is useful to teach so that students tune into the structural parallels between database interfaces. At NCSU in our first year writing instruction we often use the comparison between Summon’s interface or our catalog’s interface and Zappo’s or Amazon’s interfaces to teach how facets work, for example. We lead off with a statement such as “this search results page probably looks familiar to you…” and point out a few things about its structure.

    For years I’ve used the example of buying a plane ticket online as a parallel to understanding what an article database is with the ticket site (i.e. Expedia) as the database, the ticket itself as like an article, and paying for the ticket as the parallel to the library having purchased the content for you (I emphasize issues around the economics of information a lot in my instruction). I haven’t incorporated this example into first year instruction, but it resonates with upper level students and grad students. I haven’t used it in first year writing mostly because I teach those classes less often and it’s less relevant to how we’re working with those students compared to other ways of getting at deeper structure.

    Anyhow, I totally agree about teaching to the deeper structure and it’s always interest to think of new ways to do so. Thanks for your reflection and sharing your strategies!

    • Dani Brecher

      Hi Kim,

      Thank you for such an awesome, insightful comment! I love your ideas of drawing parallels between activities that people already engage with and databases, especially for something as complex as the economics of information.

      Your thought about if Google Scholar and discovery systems as game changers has got me thinking. One interesting thing I’ve found at my institution is that people use Google Scholar, but they hardly ever use our discovery system (I do, since we had Summon at UNC and I learned to love it there). But, overall, I would say that people still prefer the more specialized databases. I’m not sure what that means yet, but I suspect, like you, that if these new tools do change how databases work, it will be a very gradual change.

      Thanks again for reading, and hope all is well with you!

  3. Pingback: On Creativity and the Value of Other Disciplines | Rule Number One: A Library Blog

  4. Pingback: Card-Sorting LIB 100 Online Outcomes | Beyond the Lectern

  5. Pingback: Serve the Servants: Do Academic Librarians Serve Faculty or Do We Have Some “Higher” Calling? | Rule Number One: A Library Blog

  6. Pingback: A Compelling Case for Active Learning | Rule Number One: A Library Blog

  7. Pingback: On Information Literacy |

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s