Saturday, October 25, 2014

How should academic library websites change in an open access world?

In How academic libraries may change when Open Access becomes the norm  , I argued that as open access takes hold, academic libraries will increasingly focus on Expertise based services like bibliometrics, open access publishing, GIS services, Research data management and more.

The question is this, with this change in focus how should academic libraries reflect these changes in priorities on their library website? It is after all the digital front door of the library and it sends out a message on what the library is about.

A typical academic library in 2010s

Academic library websites in the 2000-2010s are all focused towards search, typically with a single search box or a tabbed search box (whether horizontally or vertical). Most sites would also have links to databases (the number of links varies) plus links to common and popular (read undergraduate) services like printing, booking of study rooms, opening hours etc.  Lastly some would devote some amount of space to news and events the library wants to promote typically in the form of some rotating banner.

I think it is fair to say this describes 90% of most academic library websites and most despite their efforts are bursting to the seams with content.

What experts think are academic libraries best practices in 2013/2014

Emily Singley currently with Harvard Library reviews academic library sites she likes, and the #1 site for her 2013 reviews is Stanford University. Which looks like this.

Noted UX expert, Aaron Schmidt was recently asked for Library Websites worth looking at.  

One academic library he went for, was MIT Libraries.

Both sites had fairly major redesigns within a year or two and spot what can be considered current state of art in academic library design. A partial hint comes from the fact that they spot dynamic opening hours widgets that dynamically state what facilities are open today, a relatively new idea compared to just static lists of opening hours on most sites now.

They devote roughly the same amount of space to the search box (though one uses a simple one box vs tabbed approach), opening hours and a news box, with the major difference being that MIT devotes a lot of screen space to displaying their research guides and experts.

In this difference hints at a possible challenge.

As academic libraries in the 2010s are adding new expertise based services and the question we face is how do we incorporate such services into the library website? How do you incorporate your shiny new Scholarly Communication team, or GIS service on your library front page? Or should you even try?

Given how political changing websites can be, easiest is to just give up and tuck the service under some hidden drop down menu or some secondary level link and hope your liaisons can make such services known.

Stanford for example tucks GIS under a drop down in "Computing - equipment & services"

Besides an argument can be made that such services are very niche (even in fairly research intensive universities like my own, the undergraduates and graduate students far out-number the research and faculty staff), so one shouldn't spend so much screen space on them.

Then again even the best liaisons can't reach everyone. And increasingly as more library effort is spent on such services it seems perverse that they don't have or have a very minimum presence on the library websites. This parallels slightly the older development where the budget on eresources increased but the manpower devoted on managing it was restricted to just one eresources librarian.

To anticipate the whole argument, I believe that while a complete redesign would be eventually necessary to showcase the new focus and provide browsing affordances, a short term solution that is increasingly adopted by leading academic librares is the bento style search box or as Lorcan Demseley calls it full library discovery

This has several advantages, one of which is it avoids a sudden shift and hence avoids the need for political battles. Users also are conditioned to search and a search box that allows you to find articles,books, AND expertise/services would scale and transition smoothly as libraries slowly (or quickly as it may) transition away from discovery/search/purchaser business to expertise business.

The age old debate on library websites

It's perhaps only a slight exaggeration to say that designing the library front page (hence forth library homepage) is a highly political affair but leaving that aside is there a "objective" way to do it?

The difficulty lies in the fact that Libraries are a mix of totally different classes of content. Traditionally the debate on library websites has always been between adding features to find "stuff" that users need to do their assignment or research (eg library databases, ebooks etc)  and using the screen space to market services/events/expertise.

The difficulty is so great, a recent 2014 article asked  Do we need two library landing pages?

The argument against Discovery on websites in 2009

But let's go back to 2009 first where Steven Bell in The Library Web Site of the Future argued that library websites should shift away from what these days we call "discovery".

He argues that academic library websites of the time pretty much fail to connect users to the full range of content library offers, as most users just use it to link to their favourite databases anyway. So libraries have failed in that task anyway despite a screen full of links and should give up trying to play that role.

He writes

"Rather than attempting to mimic search engines academic librarians should aim to differentiate their Web sites. They should devote the most eye-catching space to information that promotes the people who work at the library, the services they provide and the community activities that anchor the library’s place as the social, cultural and intellectual center of campus. That shifts the focus from content to service and from information to people."

He then asks what do we do with all the expensive databases that library buy if we don't put links on the library website?

Bell clearly anticipates some of the more common ideas currently in play in 2014. He talks about how users "invent their own backdoor routes to the content" and how we should be embedding library service so  “we’ll be where you are.” (eg Library Links in Google Scholar) or as another meme goes "Discovery happens elsewhere".

However he does propose a rather unusual idea that the LibGuides that librarians create for specialised courses and serve as a substitute for listing databases on library websites. This drew a response by Catherine Pellegrino - The Library Web Site of the Future: thanks but no thanks.

The main argument seems to be 1) The idea of providing LibGuides for every need can't scale and 2) Following Steven's idea would lead to the now famous XCD comic about the difference between what users want to find on a University Homepage and what the marketing people want to sell.

Has the argument changed in 2014?

I would argue that Steve Bell's argument for devoting less space to content for "finding stuff"  is even stronger in 2014 then in 2009 for the following reasons.

Firstly back in 2009, web scale discovery was just being launched in the first pioneer libraries. Whatever shortcomings there are in modern web scale discovery services it definitely does a job better at the exposing the full range of content a library has then 50 links to databases.

While we know that users no longer start their research from the library homepage, they do eventually return to the library homepage. Part of the reason is to check for known content, in which case our web scale discovery tools are light years ahead of the federated searches of 2009.

Because of that academic libraries of today can get away with using less space for searching features and yet be more effective.

Secondly, I have controversially speculated that as open access takes hold, Libraries roles in traditional discovery and even fulfilment/delivery for users will diminish .

As of 2014, whether libraries should bow out of discovery roles is been heavily discussed, for example see
Does Discovery Still Happen in the Library? Roles and Strategies for a Shifting Reality. One of the interesting points he makes I think is that current web scale discovery services are failing to match the Googles of the world, not due to the size of the index but the lack of "deep personalization".

 Jill O' Neill asks "Should libraries abandon investment in formal discovery services (of whatever ilk) and leave the job to the somewhat mysterious algorithmic mercies of Google, Amazon...."  Like Roger Schonfeld, she asks “Can libraries step forward to play a greater role in current awareness? Should they do so?”

For other responses see the follow up post by Roger Schonfeld

This debate is up in the air, though  Barbara Fister points out even in the day pre-web when it came to discovery academic libraries were  "supplementary, not the only way discovery happened". 

For now, I am going to be pessimistic and say No, libraries should eventually (when I do not know) prepare to walk away gracefully.

This is an argument that is not one that I make happily. However it does make the tussle over the academic library website much easier because there is one less thing off the agenda to focus on. 

That said, personalised recommenders strike me to be a totally different class of services then simple search boxes or links to databases. For example a Journaltocs type system enhanced with Google Now level smarts and perhaps mediated by knowledgable librarians strikes me as fundamentally different and something users would want assuming we can do it better than Google etc. But those don't require a search box....

The library website of the future will focus more on services and showcasing expertise

The typical academic library of 2014 now has a lot more services or put a lot more head count on them than 10 years ago and this will accelerate as the library loses it role for discovery and fulfillment.

Most typical academic libraries (granted of at least a certain research intensity), will be at the very least be working on promoting open access and populating institutional repositories, perhaps managing APC funds. Most will also be assisting with bibliometrics, benchmarking etc.

Depending on how progressive the academic library is they might be also be supporting other parts of the research life cycle.
  • GIS services
  • Data research management services
  • Grant management - playing a role in CRIS/IR integration 
  • Supporting innovative teaching environments like MOOCs

As the library shifts to devote effort to such initiatives, the academic website of the future will have to correspondingly devote more screen space to such services. Slowly there will be a shift towards using more and more space on such services and less on searching features.

Could we see a day where search boxes become de-empathized, or even disappear totally and exist merely as links as a rare few libraries do?

How much space are academic libraries devoting to search vs marketing now?

A Use of Space: The Unintended Messages of Academic Library Web Sites a study of 50 academic libraries and how they devote screen space to various categories of resources and services is probably worth looking at if you want to know what academic libraries are doing now.

Somewhat to the average space taken by marketing/PR is pretty high 28.2% (excluding white space etc) and and is ranked first.

Multisearch box comes in 2nd on average using 10.2% of used space.

The paper points out variance is high, range of space used for marketing/PR varies from 4.6% to 67.9%, with the paper heavily criticizing the instance of 67.9% usage.

According to the paper one of the most important findings is

"Occupying an average of 28.2 percent of the used space, promotion/PR is clearly important for all sites. This indicates libraries are trying to engage their users with more than just text-based communications. The message seen in this content is libraries know they have value to provide and are working proactively to connect users to these essential resources and services rather than waiting for users to locate them on their own."

Not bad at all isn't it? Academic libraries are trying to market... That said Steven Bell in The Library Web Site of the Future would argue

"It’s not that academic library Web sites completely ignore marketing. It’s just done badly. News about the library’s programs, events or new resources are often crammed into a corner of the page, are limited to small bits of text or are relegated somewhere out of the F-zone, the area, according to usability experts, to which most web users’ eyes naturally gravitate. Those prime real estate areas are instead dedicated to lists of links to catalogs, database lists and things with names that mean little to anyone other than a librarian."

So his problem is that such marketing are not prominent enough. 

Marketing - a dynamic short term effect?

I think most marketing/PR type content consist of rotating banner type ads, which markets whatever is the "favour of the month". Are they effective in pulling in eyeballs and clicks? My own personal experience at my institution is while they don't beat mass email marketing, yes they are effective (contingent on positioning on your website).

That said, I am not a big fan of such dynamic content at least for long term efforts. So your library just did a big push for "shiny project or service of the year" and you splash it or over the big rotating side banner. All well and good for a quick boost in publicity but not something you would rely for in the long term on for a core, essential part of your library offerings.

If academic libraries are going to move toward expertise based services where librarians are your greatest resource, I would advocate a permanent, prominent spot on your library homepage to advertise your library expertise and services, what MIT Libraries does for their libguides and experts is a start.

Are bento style boxes or full library discovery the (short term) answer?

One must be careful though of avoiding the problem pointed out by the famous xkcd webcomic


A redesign that goes too far , too fast before users are ready could backfire. The open access tipping point isn't easy to predict so a redesign that moves away from discovery is difficult to time.

Tons of user testing needed here.....

A possible solution is the bento style idea. I've written about the bento style idea at least half a dozen times.

Here's again the famous NCSU one that started it off.

Notice this form of bento style is distinguished from the following type of display of search results.

Lorcan Demseley called it the difference between full library discovery and full collection discovery.

The Villanova University Library example  which is common with many Universities using Vufind etc, is still focused on showing only standard results from discovery systems including books and articles.

 I have argued that this form of presentation is designed to help with relevancy issues. This is particularly useful for users who only have interest in typical catalogue results such as books or multimedia material, which may be buried in a more conventional "blended" single style result.

It also helps to cater to more sophisticated users who dislike the all in one blended approach and prefer results segregated by format.

However this still doesn't go far enough. The way ahead I agree is "full library discovery" and not just full "collection discovery".

The next generation website and also search will have a bento search that includes various sources, not just collection results, but also services, people expertise results.

This can be accomplished most simply by pulling in results from a general website search. More refined approaches would pull in anything from

  • Libguides
  • Faqs
  • Some smart matching of Librarian profiles (See Mlibrary)
  • Best bets - whatever is popular or uniquely searched by your users

Why is this important? Simply put, this allows the academic library to showcase not just the collection but also expertise and services.

So while Stanford library does not make it quite so easy to find GIS services and expertises directly by browsing, a search will surface it easily.

It is also a low risk strategy as compared to revamping the whole academic website.

I have not done a recent survey on academic libraries that use bento-style results from their search but it is likely to be popular with the biggest and most research intensive libraries.

MIT Libraries currently does not have it yet but their environment scan report prior to implementing Ebsco Discovery notes

"With the exception of BU, the sites that did offer a “Search All” option on their homepage, typically pointed users to a homegrown “composite library search” application that unified search results from multiple sources in a tabular user interface (see Stanford, NCSU, Dartmouth, Columbia, Virginia); with the exception of Stanford, these sites utilized the discovery service as a data source (via API) to populate an articles/e-books section of the composite library search results" and plans for one in the future.

Yale University is also implementing a quicksearch that is supposedly inspired by the Columbia example., though the current example includes only collection based boxes.

What are the drawbacks of a bento style box?

The most obvious drawback is users are used to Google and Google Scholar style results.

Here's one way to put it.

"At present, we are not using bento. This is because our community is familiar with Ebsco's blended interface, in which the results are presented as a list, ranked by relevancy based on the available metadata."

From here to discovery

Some argue Google, does some degree of segregation, though I would argue not to the level Mlibrary or NCSU (another interesting example) does it.

Recently work by Karen Mills of the Open University What do students want from library discovery tools?
suggests that students are confused by Bento style results. This is probably due to confusion with the labels and terminology.

I suspect, the effectiveness of bento style boxes would be affected by the number of boxes and the type of users. Freshman with almost no academic experience will probably dislike bento style. On the other extreme, a experienced graduate student or faculty with good mental models of research would appreciate a bento style display especially if they are predisposed to utilize the other non-collection based services (eg consulting of librarians) that such a search would throw up.

As such, I would expect the most research intensive University Libraries who have a higher share of such advanced users to push for bento style searches, while community colleges might go for a blended google style results display. This is what we see of course, though this could be driven by the fact that bento style display results more resources to implement since currently the major web scale discovery services do not provide bento style results out of the box.

A personalised approach that allows one to pick either display is possible of course.


Let me end by quoting Bell again.
"Academic libraries must promote their human side. The library portal experience should emphasize the value of and invite stronger relationships with faculty and students. That means going beyond offering a commodity that, by and large, the user community can well access without the Web site. The next generation academic library Web site must leverage what academic librarians can do to help faculty and students improve their productivity and achieve success."

With the rise of open access, the above has never been more true.

How would such a website look like, when the academic library is freed from the responsibility or duty of discovery? In the long term, I believe it should be radically different but that is for the future say 2020ish. For instance how would a "people/expert first search" look like? Would it simply be a CRIS system like Vivo?

blog comments powered by Disqus

Share this!

Creative Commons License
This work is licensed under a Creative Commons Attribution-ShareAlike 3.0 Unported License.
Related Posts Plugin for WordPress, Blogger...