Ranting about library technology

I’m meeting with my Dean later today to discuss the libraries’ strategic plan with relation to technology. I’ve been thinking about this a lot over the course of this morning and afternoon, and I thought I’d share a bit of a rant about library technology, and things I think need to change in the future.

1.Multiple systems, or at the very least, the appearance of multiple systems, are enemies to usability.

I recently had a conversation with my colleague Lynell. We were discussing library tutorials, and criteria to use in deciding when a process should be explained in a screencast versus and when a process should be explained in simpler media (with text & screenshots). One thing Lynell recommended is that in making this decision, I should consider the number of systems the user needs to cross.

For example, I’m working on a script for a screencast tutorial on citation chasing. Our user starts with the library home page (our CMS acts as system #1), then searches our A-Z journal list (system #2) and hopefully is then taken to a database which contains the full text (system #3a). Of course, that database might not always work (WILSON SELECT PLUS, I’m looking at you) which means the user then has to go back to system #2 and then pick system #4a. If we don’t have the full text, the user then has to search the catalog to see if we have the journal in print (system #3b) and if we don’t, they need to use document delivery to request it (system #4b).

That strikes me as needlessly complicated. Our users don’t need to know that they’re traversing from the CMS pages to a journal list to FirstSearch. They don’t need to know that our document delivery system is called Illiad. They just need the blasted article. Furthermore, the more systems the user must traverse, the more complicated troubleshooting becomes for library staff. Few of our users are sophisticated enough searchers that they know why their searches failed somewhere along the line. They only know they had technical issues. I’m sure a lot of time they don’t even know that – or if they do, they just give up and try something else.

2. As a corollary to my first statement, indexing databases are a bane to usability.

In my past life, I was a reference librarian. I did instruction and liaised with several departments. I certainly understand the need for subject-specific resources. I don’t advise abolishing them. Unified discovery systems can’t do everything – I mentioned Summon to a group of senior-level geology students last semester, but I herded them toward GeoRef to find articles, and Web of Science to determine the impact of those articles. Summon would have brought too much “noise” into the equation.

All of that said…resources like Philosopher’s Index get my dander up, simply because they add another system into the equation.

Furthermore, the fact that most vendors that provide seamless links to full text resources in other databases makes this situation more complex. For example, America: History & Life is a resource my library subscribes to through Ebsco. Some of the citations that come up when I search, then, have full text – because they’re linked to the full text we have in other Ebsco databases like Academic Search Premier. One one hand, I really like this, because it means fewer systems for the user to traverse. On the other hand, the citations that have full text are not necessarily the best citations. Also, it means I may have to explain to users why some articles have the full text and others don’t. Do you think they care? I don’t think so. They want what’s there.

As I mentioned above, I was a reference librarian in my past life – and I want to help our users find the best resources for their information needs. But I know that they’re going to go for the resources that are easiest to get their hands on. So let’s try to work toward making the best resources easier to get to, hmm?

3. Vendor branding is at odds with the mission of most libraries.

Going back to my first point – we at least need the appearance of fewer systems. This means that we need more power to customize. More power to make systems look like one another.

Going back to my second point – I want to help lead our users toward the best resources. Vendor branding seems at odds with this goal – especially since it’s vendor branding and not resource branding. If I threw a screenshot of a search screen for Academic Search Premier up on fivesecondtest, what do you think people would notice first, the prominent blue EbscoHost icon, or the bold print that reads Academic Search Premier? This isn’t helpful to our users who want to go back to the resource they used before.

Unfortunately, I’ve also seen that our users become used to using a particular resource because it’s memorable – because the vendor branding has worked. Unfortunately, again, the “easy” resource isn’t always the best resource.

JSTOR’s a good example of this. I suspect, knowing what I’ve heard from friends and users,  that it’s one database that student use for everything, because they don’t know we have other resources that might be better. In the LibGuides survey we conducted back in 2008, one of the questions we asked of our users was whether they preferred links to databases or links with descriptions of the database. I used a few screenshots of links to JSTOR as an example. My description of JSTOR, which I’d written for a stats guide, I think, read, “JSTOR is a multi-disciplinary database, but has some scholarly literature in mathematics and statistics. Be warned, though – JSTOR is an archive and does not contain articles published in the last 5 years.” Although we learned that our students do want database descriptions, we also learned that a lot of people didn’t know that JSTOR didn’t contain recent articles (via the comments). They knew JSTOR existed, but not what it was.

They think JSTOR, they EBSCO; they don’t think library. That’s not going to help them when they graduate and they don’t have access to these resources anymore.

I see why it’s in the vendors’ best interest to make themselves visible and memorable, but it’s not helping our users get to the best information, and I have a problem with that.

4. Metadata needs a makeover.

One of the reasons I think unified discovery systems work (well, okay, why Summon seems to work – I can’t comment on others) is because the search engine is working off of a single index which at least has some consistency in how metadata is presented and searched. I think one of the reasons federated search doesn’t work is because the metadata is coming from so many different sources that it just can’t be translated consistently. It seems a huge waste of time that vendors have people working on connectors to read that metadata and parse it. Why can’t we just create the metadata in a consistent fashion in the first place? OpenURL is supposed to help with this, if I understand correctly, but it doesn’t seem like enough. I honestly don’t know what the faults and pitfalls of MARC and AACR2 are, but I do know that it doesn’t matter whether I’m searching Worldcat.org or Millennium or an Evergreen system – if I search for “Cookery for one” as a subject, I will find books on cooking for one. I have no such luck with articles. Furthermore, I don’t know who the heck is indexing these articles. It seems to me that most of WorldCat is user-contributed records, right? I have more faith in something that’s crowd-sourced than I do in something that’s vendor-sourced.

I’d also like to see some usability studies on what metadata is actually useful for users (tables of contents? book covers? etc.) or where a lack of metadata stands in the way of users finding good resources.

I’m not really sure what plan of action I can take to accomplish these goals, at least at this point. I’m not even sure if these are things other people think are problems. Do you agree with me? Do you disagree? Care to elucidate strange inner workings of Library Land that I do not understand?