The dumbing down of the library tools While others seem to be celebrating the googlizing of the OPACs and databases, I have some strong reservations.

According to the database providers, end users want to find a box wherever they are (in e-mail, in word processing software, on the desktop, in a browser), put in a maximum of two words and have exactly what they need from all collections of information (the web, ejournals, books, etc.). CSA's upcoming release is a step in that direction, but what you get is more like mystery meat than the best answers from the correct source selected for you. Slight disclaimer here: it's not actually been demonstrated for me, these are my observations from playing around. For the trial it starts you on the easy search. We have the choice of subject areas: "Natural Sciences" or "Technology." Hmm. Those are real specific. It's yet another click to select which database you want.

Let's try a real world search. hypersonic AND isolator*. I know that this is in A&HT. 0 results under technology. 0 results under natural sciences -- wait , wait -- don't tell me that oceanic abstracts is the default database for technology?! Ok, I get 22 results when I select A&HT (and I get 17 in the old interface?).... ok, so now I got out of the database and got back in and now "technology" gives me the same 22? Operator error? Don't get me wrong, once you get the results the suggested terms are really cool. I also like being able to add them to the search with either AND or OR.

We also were just shown the Jan 2005 release of EV2/Compendex and it has a new easy search, too. With EV2 the library can decide what search screen the users will start with so tentatively we can just ignore the easy search like we do with Ebsco. This might be the case with the CSA product- I don't know. Another potential difference is that EV2 recommends limiters based on your search. I can see a customer actually using this as a shortcut instead of playing with the thesaurus. In general, though, my customers are very savvy information consumers and I don't think it's fair to assume they can't grasp more complicated search screens.

That really sums up my problem: web searches get hundreds of thousands of results and we know that users will seldom go past the first page or so of results. When databases have just this one block-every field in the database search and then sort by relevance... what confidence do we have that the user will find anything of interest? What if it sorts by date? You might have something completely bizarre come up first. What if you're an electronics engineer searching in the CSA easy search for "translator" and all you get are things from the polisci and literature databases when all you wanted were frequency converters! Do you call the library or just assume that there's nothing there for you? (and then go around grumbling about how the library doesn't serve engineers).

Ok, so what's my point? In a research setting like a university or research lab, offering the easy search might be attractive, but doesn't in the end serve the customers well. They will be tempted to use it and will not get the best results out of the databases. IMHO.
  ¶ 9:14 AM| |cites (technorati) |