Tuesday, August 26, 2008

Your local library

So, apparently faculty have changing views of their campus libraries.

Since some of you probably can't access the article, here's the salient point: "However, it also confirms that researchers increasingly find what they need through Google Scholar and other online resources, a trend the report's authors anticipate will accelerate as more and more knowledge goes digital. ... But only 48 percent of economists and 50 percent of scientists value libraries as gateways."

There is something wrong with this view. Google Scholar allows you to find articles, but - and this is a very important but - not to see most of said articles unless one is connected through an institution that has paid for access to those sources. So when I hop on to Google Scholar to find articles on lemur femurs, it's the VU library that gets me there. And I'm very aware of this because after many of the results it shows "findit@VU." The physical library may not be a gateway, but the virtual library is.

Are other scholars really too thick to notice this, as the piece seems to suggest? Or are they simply not relying on articles? Some scholars may be taking the easy route. I know students who won't bother to use anything that is on microfiche. From reading dissertations, I know a lot of them don't bother to scan other dissertations first. But many economists increasingly publish working papers online themselves long before they appear in journals. They may, quite validly, be sidestepping the library. Yet I don't think that you can completely bypass the library in any field.


Rebecca said...

A few thoughts:

- I agree; google scholar doesn't let you get at most of the articles you see. However a carefully crafted search could point you to the items you need more easily than some of the search engines that schools provide. Too many different subscription service search engines, compared to a single google interface.

- I also know of some people who will use google scholar to get a citation without ever reading the whole article. A couple of sentences or someone else's reference to something can occasionally provide enough (although there is no guarantee that they will accurately reflect the original article.)

- On the other hand, I am finding google books invaluable; Historical books are often completely available and you can usually see the relevant pages of newer books if you need just a snapshot.

- I never look at old dissertations. In MOST cases the person has published their findings at least once or twice in a peer-reviewed journal somewhere, which seems to have more credibility and is always shorter.

turducken said...

I certainly don't want to beat on Google Scholar; it's awesome. But it's really only powerful in conjunction with a library.

The "cite without reading" thing is such sloppy scholarship. If you're using someone else's cite, the appropriate thing to do is cite "Smith, as discussed in Jones (1973)." Otherwise, as you say, you have no guarantee you are accurately getting at the original piece. It's also how errors get propagated virally - when you see something mis-cited in several sources the same way, you know who hasn't been doing their reading! Sure, people do it.

The dissertation thing varies by subfield. My adviser and I have been working on a lit review of dissertations about higher ed fundraising. 91% of them never reappeared in any other form. I've seen similarly high numbers for other practitioner-dominated fields such as public administration. Obviously, if it's not the case in your area, one can reasonably assume the few dissertations that aren't published probably aren't very good. But if there's a high non-publication rate, you get this ridiculous vicious cycle where everyone says, "There's no work, therefore mine is needed," but they never publish it, and so the next person says, "There's no work .."