You don’t have to be an expert on user experience to conduct your own usability study. Determining how your users prefer to use your site, collection, product, etc is something that everyone can benefit from and anyone can do. Usability studies are conducted to understand how your users actually use your system. You can then use that data to determine what is and is not working with your system and create an improved system that has greater usability.
We are currently conducting a usability study at my library. We provide research aids using a content creation software called Libguides. Our goal is to provide Libguides that students will have the best experience using. One reason I saw a need to change the design of our current Libguides was due to a steep drop in clicks from the Libguide homepage to subsequent pages. Something about the way the guides were laid out was causing users to navigate away from the page. So with the help of our systems librarian we set up a usability study to discover the reason. We have only gone through two tests so far, but have come across some interesting findings.
- Usability Test 1:
- Students were asked to draw their ideal Libguide on a large piece of paper
- The systems librarian and I explained what Libguides were and showed them Mount Holyoke, UT Arlington, and UMHB Libguides as a reference
- Usability Test 2
- 3 prototypes were created based on student suggestions from usability test 1
- Prototype A: Based on Mount Holyoke’s Libguide layout
- Prototype B: Based on UT Arlington’s Libguide layout
- Prototype: UMHB’s Libguide
- All dealt w/ the subject of education
- After the prototypes were created a series of tasks were composed
- 4 users were given the tasks and asked to complete them for each prototype
Below is a list of our findings and recommendations for our next test.
Remove Top Level Navigation
- None of the users used the tabs at the top. According to the literature users do not utilize top level navigation, favoring left side navigation instead. Below is an example of what we currently have and what I would like to migrate to:
Current Layout: In addition to top level navigation, there are several extraneous links that navigated away from the Libguide and confused the user.
Desired Layout: This is taken from http://www.fall.tnvacation.com. I like how the active tab is a different color, as well as the clear intuitive left side navigation.
More White Space
- Students did not like text heavy pages or pages with too much content.
Current Layout: Too much text.
Desired Layout: This Libgude is from UT Arlington. They employ a good use of white space as well as images.
Less Database Listing, Incorporate Search Boxes
- Students became overwhelmed by extensive listings of databases. They preferred shorter lists and embedded search boxes.
- Here are a few more recommendations I am making for the next stage. In the coming months we are going to create a new prototype and test a new group of students on it.
- Put most important links at the top of a list
- It was observed that users do not look at the link titles but rather click on the first link/search box they come to
- The literature reports that students will often search any search box available- regardless of its intended use
- Have all links open in a new tab/window when clicked
- Be clear w/ terminology
- Don’t use library jargon
- Students did not know the difference between database and article terminology or EBSCO and specific database titles
- Tab titles should have titles that are easily understood (finding articles, finding books, etc)
- Need to utilize repeat navigation
- Users should be able to easily navigate back to the home page of the libguide
- Libguides should focus on:
- How to use the library catalog
- Keyword and subject searching
- Understanding what a database is and how to use it
- Choosing a topic for research
- Mind mapping
- How to differentiate between scholarly, popular, trade publications
- Find permalinks, full text, etc
- Find useful Websites associations, etc
- Libguides should NOT focus on
- Extensive listings of books we have on a subject (it should show HOW to find those books)
- Purpose of Libguides should be better explained
- Students did not know the difference between the library site and the libguides
- Once they navigated away from the libguide they rarely navigated back to it
- Literature reports that lack of clear navigation through the site leads to disorientation
10 thoughts on “My Library Usability Study Stage 1”
I am a HUGE stickler for good UX/UI. I’m so picky nowadays, I won’t even use sites that I don’t find easy to navigate or aesthetically pleasing.
Me either! I have started using the Webby award winners for best user experience as inspiration for future designs. http://winners.webbyawards.com/2013/web/website-features-and-design/best-user-experience
I clicked on this from AL Direct. This will help when we begin reorganizing our library’s webpage(s). Very concise. Also, I’ll have to take a look at the Webby Awards that Book Maven mentioned. Thanks!
No problem glad it was helpful!
Lots of food for thought here, and more than a little for debate; but too much to address in comments to a blog post. I think the “Other Findings” are particularly useful, but they do raise questions: (1) at what point does a list become “extensive” or text become “too much”? (2) how much of #7 could and should have been covered in the primary and secondary grades so students would know these things before they hit college?
Unless their campus offers a full semester-length course in information literacy, most academic librarians don’t have near the student contact hours to address all these things, yet we are finding that we cannot count on K-12 to have provided any of it. It’s like trying to build the second story of a house when the foundation has not been put in place.
Yes the study does definitely raise some questions. The team I am working with on this project are trying to decide what constitutes as “too much” text. We are going to make a new libguide, with less text and more white space and see how it tests. One thing we want to do is make sure the most important information is available without having to scroll down (“above the fold” http://en.wikipedia.org/wiki/Above_the_fold). I think most of the objectives of #7 should be covered in primary and secondary grades and continued to be emphasized on the college level, but like you said that is extremely difficult. Right now all we are able to offer is a “one-shot” information literacy session that is not enough time to teach the concepts needed for proper IL. I am looking into “flipping the classroom” so I can use my one hour class time for discipline specific hands on activities. In a flipped class students will be required to watch a library 101 video prior to class, so we don’t take valuable class time to teach basic library information. It’s not ideal, but I think it’s a step toward a full semester-length IL session. I want to assess students searching improvement before and after taking the one shot sessions, present it to faculty at our faculty assembly, and gain more faculty buy in. I have found that people don’t argue with hard data. Does your school program allow you to offer library orientation sessions?
[ Okay; I don’t know if this reply is going to your personal e-mail or to your blog, but
here goes. Since you went long, i won’t worry too much about brevity and, to minimize confusion (since I cannot change font colors), I will put my comments in brackets.]
Yes the study does definitely raise some questions. The team I am working with on this project are trying to decide what constitutes as “too much” text. We are going to make a new libguide, with less text and more white space and see how it tests.
[My first thought when I read your initial post and saw all those comments about “minimal text” and “no extensive lists” was, “And here am I, the chiefest of sinners.”
My workhorse LibGuide, the one I had to create as it replaces a “Databases by
Subject” page that was going to vanish into the cybervoid, reeks of text (minimal white space), and its three top level-indexed pages are (alphabetized) lists of databases. All these errors being admitted, that LibGuide still has recorded around 3300 hits in six months, even when classes are not in session.
My other, upper level course-anchored, LibGuides are by no means impoverished of text, and they also have lists of circulating and reference books. The idea is to get brick-and-mortar students into the stacks where they can take advantage of the serendipitously browesable Library of Congress classification system. The system is by no means perfect, but it does appear to bear good fruit.]
One thing we want to do is make sure the most important information is available without having to scroll down ().
[Even if we could accurately predict which information is most important for every user, we still face the “Goldilocks dilemma”: when does too little information become too much information as we strive to reach “just right”? More text means more scrolling. A presenter at the 2013 Alabama Library Association conference elaborated on this with reference to distance education students. Absent the possibility of face-to-face dialog, LibGuides are the best medium for conveying information, unsatisfactory as they are. But aside from substituting bulleted lists for paragraphs, how are we to reduce text and increase white space?]
I think most of the objectives of #7 should be covered in primary and secondary grades and continued to be emphasized on the college level, but like you said that is extremely difficult. Right now all we are able to offer is a “one-shot” information literacy session that is not enough time to teach the concepts needed for proper IL.
[I believe you are in Texas, whose legislature chose to defund its K-12 package of academic databases some years ago. This forces Texas students to travel to the land of Binglehoo to practice their information-seeking skills. Alabama and Georgia still have their database packages, but I have far more evidence that they are not being used in K-12 than that they are. Thus the students’ reliance on the internet, a habit higher ed has a tremendous problem weaning them from.
In my opinion, the situation described above forces ILI librarians in higher ed to continuously work from the wrong end of the inductive-deductive reasoning chain: we have to teach specific databases (and the catalog) in the context of specific assignments. That is what the discipline faculty who request the instruction sessions expect. But where do any of us have the opportunity to teach the broad transferable skills like how to use databaseS: related terms, truncation, doing background to select a topic and then refining it? The difference in document types and formats? If students could get these things starting in grade seven and increasing in complexity through grade twelve, then the one-shot lecture just MIGHT be enough in college.]
I am looking into “flipping the classroom” so I can use my one hour class time for discipline specific hands on activities. In a flipped class students will be required to watch a library 101 video prior to class, so we don’t take valuable class time to teach basic library information. It’s not ideal, but I think it’s a step toward a full semester-length IL session. I want to assess students searching improvement before and after taking the one shot sessions, present it to faculty at our faculty assembly, and gain more faculty buy in. I have found that people don’t argue with hard data. Does your school program allow you to offer library orientation sessions?
[My library has eight public services librarians, and we all offer instruction sessions in our disciplines. Your before/after comparison is as good a way as any to get the ball rolling for your project. You may wish to contact the librarians at Samford University, as they have some experience with these sorts of comparisons.
As to the semester long course, the University of West Georgia has had one for
fifteen years, and a campus (whose name escapes me) in the University of Missouri system also has one. If you want to compile a roster, a Google search
(because I was too lazy to switch over to Dogpile) of universities with “information literacy course” might do you some good.
Thank you for your very insightful comments/input. In our testing we noticed that although a Libguide had a considerable number of hits, statistics showed that page views to subsequent pages on that guide decreased considerably. Also the “clicks” recorded for each link on the page was not as high as we expected. Clicks also decreased gradually from the first to the last, indicating that users tend to click on the first link and not links below it. We want to utilize Google Analytics in the future and record how much time is spent on each page and how often users scroll down. If users are only viewing the page for a few seconds then we don’t see it being as useful as we’d like. I just returned from ALA and in a User Experience panel I attended librarians from the University of West Georgia reported that students didn’t know what the list of databases meant on Libguides and preferred more images and research process explanation than text and lists. But every student population is different. Your students may utilize database listings more than ours or other libraries. I think that’s why it’s important to study user behavior in specific settings because student populations are so diverse. And like you said finding how much text is “just right” is a dilemma and one that we are currently analyzing. We reference W3C standards on Web Design as much as possible http://www.w3schools.com/web/web_design.asp to determine what we should and should not use/include on our guides.
It is also our goal to make Libguides an enjoyable experience for the user. The literature states that a lack of problems is not the same as a positive user experience (Hassenzahl). We want to create a pragmatic and hedonic Libguide experience (Hassenzahl), and our user studies indicate that our student population are not getting that.
At ALA I viewed SpringShare’s demo of their new UI. I think white space is going to be much more achievable with the new layout. Boxes are going to have only three format options and are going to allow multiple content types in the same box (one box can have video, links, RSS feeds, etc). Having to create a new box for every content type has added to the cluttered view of the Libguide pages. I’m really excited to start utilizing their new interface this Fall. I feel like with a more user friendly layout we can focus more on what content we want rather than what design we should have.
Yes in Texas, because students don’t have a strong background in database searching we have a big problem weaning students off Google into more sophisticated tools. That’s a really good point which I will present to the faculty. I think it emphasizes the need for an semester long IL course, which is what we’d ultimately like to do. I will definitely contact librarians at Samford and ask for some suggestions/best practices thank you for the recommendation! Right now we are curating a list of quality IL videos from the Internet to use this Fall.
If nothing else, I suppose our dialog proves that what may be the best mousetrap
for one demographic may not be at all suitable for others. Our analytics do not
appear to be anywhere near as sophisticated as what your are using — I think we
just have the standard Springshare package — but my LibGuides (cannot speak for my colleagues) do not mirror your experiences. My single-purpose workhorse guide that gets so many hits has its hierarchy in the sequence of pages (tabs). The homepage is designed for the freshman English students, who really have no need to navigate to the succeeding pages. Those pages are more for upperclass majors and faculty and, understandably and by design, receive fewer hits.
Having seen LibGuides from several university campuses that have so many pages
that the top-level indexing looks like a forest of tabs, I try to keep my LGs to no more than six pages and maintain a logical sequence among those six, simply so users won’t be overwhelmed by options after they hit the homepage. Sometimes they bypass the homepage completely and go to the page they want.
I wish I could have attended that User Experience panel. I did look at several UWG
LibGuides, and I kind of see what their students may be driving at. As might be expected, style and format vary by librarian. I noticed some LibGuides that are text heavy (like mine), but there are several that have unadorned lists of links (not many widgets that I saw) with no accompanying explanation. I can understand students’ unhappiness with that, since it infers they have a body of prior knowledge which probably they do not. I cannot imagine what sort of images the students have in mind any more than I can see a way to explain the research process without text. I did see some things on UWG’s LibGuides which will help me with my own.
Since you mentioned SpringShare’s new UI, I too am curious about it. I just hope that it gives us more without taking away something, which happens all too often IMO with updates and upgrades. I approach the internet searching dilemma by pointing out that if a thing is not a webpage itself, or linked to a webpage, it cannot be culled via an internet search. Google Scholar is sparsely populated and much of what it returns in its results lists as scholarly, just isn’t. More things are becoming available through Open Access, but not nearly enough yet for that to become the first search option. Students (or their parents) are paying big bucks to pick the minds of professors, and those professors do not send their research to the internet for publication; it goes to the information sources bundled in the databases. Which students are reluctant to search. An irony which continues to baffle me.
I see I’ve gone long-winded (again). I have enjoyed and profited by our discussion, but probably I should let you get back to your regularly-scheduled life. If nothing else, our exchange has motivated me to see if I can streamline the text of my LibGuides down to just the essentials.