All posts by Angela Zoss

Where do patrons get lost? A study of library navigation.

In the Assessment & User Experience Department, we’re always looking for ways to improve how our patrons experience the libraries’ physical and online spaces. One of our primary ways of learning about our patrons is our biennial user satisfaction survey, which we use to collect opinions from large groups of our patrons about a wide range of issues.

One topic that comes up regularly among our patrons is the navigation of our physical spaces. Like many libraries, our buildings have evolved over time, and that can make navigating our spaces a bit complicated. On Duke’s West Campus, we have three library buildings that are interconnected – Rubenstein Library, Perkins Library, and Bostock Library. Responses and comments on our biennial survey confirm what we hear anecdotally – patrons have trouble navigating these three buildings.

Deep Dive into Navigation Concerns

But how can we follow up on these reports to improve navigation in our spaces? Ideally, we would gather data from a large number of people over a long period of time to find very common and problematic navigation issues. Our biennial survey offers data from a large number of people over time, but it isn’t a great format for gathering detailed data about narrow subjects like navigation.  Conducting an observational study of our spaces would explore navigation directly, but it would only include a small number of people, and the likelihood that we would catch individuals having trouble with navigation is low. We could try conducting ad hoc surveys of patrons in our spaces, but it would be difficult to ensure we are including people who have had navigation trouble, and it may be difficult for patrons to recall their navigation trouble on the spot.

What we needed was a way of capturing common examples of patrons having trouble with navigation. We decided that, instead of asking patrons themselves, our best resource would be library staff. We know that when patrons are lost in our buildings, they may reach out to staff members they see nearby. By surveying staff instead of patrons, we take advantage of staff who know the buildings well and who are commonly in particular areas of the buildings, noticing and offering help to our struggling patrons.

We decided to send a very simple survey to all staff in these library buildings. Staff could fill it out multiple times, and the only two questions were:

  • What is a common question you have helped patrons with?
  • Where are the patrons when they have this question, typically?

We had a great response from staff (72 responses from 36 individuals), and analyzing the responses showed several sources of confusion.

Pie chart showing categories of responses to library navigation survey. 35% of responses reported patrons in the wrong building. 26% reported patrons in the right general area. The rest were split amongst "wrong campus building," "wrong floor", and "can't find books."

Focusing on questions where patrons are in library spaces and not near a help desk, two concerns account for over 60% of reported patron navigation issues:

  • Building confusion
  • Hidden rooms

Trouble Between Buildings

Our three connected library buildings, unfortunately, connect in ways that are not obvious to new visitors. Because buildings only connect on certain levels, it is easy for patrons to be looking for a location on the right floor but the wrong building. By asking staff for specific locations of both patrons and their desired destination, we could compile the most frequent problems that involve being in the wrong building. Unsurprisingly, the locations that cause the most difficulty are our large meeting rooms and classroom spaces, especially those that are not on the ground floor of the buildings.

The most common problems seem to happen when patrons leave the first floor while in the wrong building, expecting the buildings to connect on the other floors (or not realizing which building they are in). As you can see from the side-view of our buildings below, the Perkins and Bostock library building have easy connections on all floors, but the Rubenstein Library only connects to Perkins on the first floor. Our survey confirmed that this causes many issues for patrons looking for 2nd floor or Lower Level meeting rooms in Perkins and upper level meeting rooms in Rubenstein.

A diagram showing a side-view of navigation between the three library buildings. Rubenstein and Perkins connect on the first floor. Perkins and Bostock connect on all but the first floor.

While we are still in the process of developing and testing possible solutions, we hope to redesign signage in a way that better signposts when patrons should proceed onward on the current floor and when they should transition up or down.

Trouble on the Same Floor

Our survey suggests, unfortunately, that it is not enough to get patrons to the correct floor. Depending on the route the patron takes, there are still common destinations that are difficult to see from stairwells, elevators, and main hallways. Again, this difficulty tends to arise when patrons are looking for meeting rooms. This makes sense, as events held in our meeting rooms can attract patrons who have not yet been to our buildings.

Staff reports for same-floor confusion focus largely on floors where room entrances are hidden in recesses or around corners and where rooms are spaced apart such that it is hard to simply follow room number signage. As a notable example, the 2nd floor of Perkins Library seems especially confusing to patrons, with many different types of destinations, few of which are visible from main entrances and hallways. In the diagram below, you can see some of the main places patrons get lost, indicating a need for better signage visible from these locations. (Pink question marks indicate the lost patrons. Red arrowheads indicate the desired destinations.)

A floor plan of Perkins 2nd floor, with curved arrows showing reports of patrons who are far away from their desired destination.

As we develop solutions to highlight locations of hidden rooms, we are considering options like large vinyl lettering or perpendicular corridor signs that alert people to rooms around corners.

Final Thoughts

This technique worked really well for this informal study – it gave us a great place to start exploring new design solutions, and we can be more proactive about testing new navigation signage before we make permanent changes. Thanks for your great information, DUL staff!

Data Sciencing our Journal Subscriptions

The ongoing tensions between academic institutions and publishers have been escalating the last few months, but those tensions have existed for many years. The term “Big Deal” has been coined to describe a long-standing, industry-wide practice of journal bundling that forces libraries to subscribe to unwanted and unneeded publications rather than paying more for a limited number of individual subscriptions. This is a practice you see in other industries – for example, cable packages that provide hundreds of channels, even if you only want one or two specific channels.

What is especially problematic in higher education is that academics produce and review the content that gets published in the journals (for free), and then the universities have to pay the publishers a subscription fee to access the content. Imagine if YouTube required a subscription fee to watch any videos, including the ones you had posted. It’s a system that makes research harder to access and inhibits global scientific progress, all so publishers can earn an enormous profit margin.

Right now, academic publishing is controlled by five publishers (the “Big Five”) – a monopoly that makes it very difficult for libraries to negotiate better deals. Only very large organizations or consortia, like the University of California, have been able to start pushing back against the system. It will likely take large shake-ups like this for any large changes to take hold, but it in the meantime there may be ways to situate ourselves for making better purchasing decisions.

At Duke, we often review our usage of specific journal titles as we prepare to make purchasing decisions. Usage data comes in a variety of forms, but the most popular are counts of Duke views and downloads that come directly from the publishers and the number of times Duke authors publish in or cite a particular journal. There are many other kinds of data that might be of interest, however, including Duke participation on editorial boards, usage differences across disciplines, and even whether or not the journal is fully open access. Blending various data sources and optimizing the search decisions for a given budget cycle can be overwhelming.

Last fall, Duke University Libraries decided to propose a project for Duke’s Data+ summer program – a summer research experience in data science for undergraduate students. Our project, “Breaking the Bundle: Analyzing Duke’s Journal Subscriptions“, focuses on Duke’s subscriptions to journals published by Elsevier. The program is in its third week, and our team of two incredibly-sharp undergraduates has been hard at work building and blending our datasets. Our goal by the end of summer is to have a proof-of-concept dashboard that lets collection managers adjust the weights of various usage measures to generate an ideal collection of journals for a particular budget.

It is still very early in the process, but the students have been hard at work and have made great progress. We decided it would be best to develop the analysis software and dashboard using R, a statistical computing project with a rich history and many helpful development tools. In addition to publisher-provided views and downloads, the students have been able to use websites and APIs to collect data on journal open access status, editorial boards, numbers of publications, and numbers of citations. All Data+ teams present publicly on the projects twice during the summer, and we hope to schedule a third talk for a library audience before the end of the program on August 2.

Sample R code from the project
Just one of many files of R code generated for the project so far.

We look forward to seeing what the summer will bring! While this project is just one small step, automating the collection and analysis of journal usage will position us well, both for responsible purchases and for a hopefully-changing publishing landscape.

Wrangling Messy Data with Airtable

The Assessment & User Experience department at Duke University Libraries keeps the libraries’ physical and virtual spaces responsive to user needs by constantly gathering feedback. In additional to our biennial user satisfaction survey, we run usability tests, hold focus groups, and host meetings of our student advisory boards, all in an effort to keep a finger on the pulse of the DUL patrons.

These activities can generate a lot of unstructured data! For example, in a typical meeting of our undergraduate advisory board, we might collect feedback from a dozen or more students, generating seven or more pages of notes and covering a range of topics. We review and act upon some of these comments immediately, but others may influence longer-term planning. As library staff, we know how important it is to store information in a way that promotes future access. This year we decided to pilot a new system for storing and describing our unstructured data.

Airtable logoEnter Airtable. If you’re not familiar, Airtable is a cloud-based database solution. Similar to Google Sheets, Airtable lets you enter and share data in your web browser, but it also offers more powerful features for projects that have messy data or interconnected components. There are many Airtable templates to show off the different features, including project trackers, event planners, and even product catalogs.

For our messy data, we built a simple spreadsheet that was general enough to collect data from a variety of sources. We included columns like basic demographics, the feedback provided, the original question or prompt, the date when feedback was provided, and how we collected the feedback. Then we took advantage of Airtable’s special features to create a column for topical tags. One of the column types in Airtable is called “multiple select“, which means you can add multiple tags to a single comment. Other spreadsheets can’t understand a list of tags in a single cell, but Airtable treats each tag separately and allows us to group and filter comments by each individual tag.

a screenshot of an Airtable database with tagged comments

The ability to look at comments across different feedback channels in one central location has enormous potential. Instead of having to hunt through old Word documents or emails, we have a single database that can be searched, sorted, or filtered to explore trends in comments over time. When a question comes up about how patrons feels about a particular service or space, we can compile data much more easily, and we no longer have to rely on our memory of what feedback we’ve received and when.

screenshot of Airtable tutorial instructions

Airtable’s free accounts have a limited number of rows allowed in each database,  but they do offer a discount on paid plans to educational institutions. We’re only just starting to explore the potential of Airtable, but so far we’ve been happy with the ability to collect our messy data in one place and organize comments with tags.

Want to learn more? Take a look at our recent tutorial on using Airtable for coding survey data, originally offered at the Designing For Digital 2019 conference.

Find your haven at Oasis Perkins

(Thanks to Assessment and User Experience Intern Brenda Yang for this post and for her amazing work on Oasis Perkins!)

What if it was possible to unwind – color, do a jigsaw puzzle, meditate – without leaving the Libraries?

It is at Oasis Perkins! This high-ceilinged refuge is tucked into the fourth floor of Perkins in room 418. It’s a perfect place to escape any finals-related tension palpable in study spaces this time of year.

A floor plan showing the location of Oasis Perkins on the 4th Floor of Perkins LibraryYou’ll find:

  • Yoga mats and meditation cushions
  • A jigsaw puzzle table
  • Coloring books, logic puzzles, and sudoku pages
  • Origami paper and instruction books
  • A quiet nook
  • A white noise machine
  • Plenty of natural light (during the day)
  • And more to explore!

Unlike other fourth floor spaces in Perkins and Bostock meant for silent study, feel free to chat and connect with a friend or strangers, or simply sit and reflect quietly.

How did Oasis Perkins come to be?

A photo of different types of teas from the Tea-laxation event.

One major motivation was direct feedback from students. Comments from our 2018 Student Library Satisfaction Survey made clear that while study spaces at the Duke Libraries are a keystone of many students’ academic lives, it can be a stressful place, especially during the exam season: “I love coming to the library during most of the semester… particularly during finals, there is an overwhelming sense of stress that emanates from the other students at the library.” A few students explicitly requested “a room to relax,” a place to have have a “refreshing study break without leaving the library somehow,” or a “stress-relief room.”

We hope that Oasis Perkins can serve as a dedicated place for students to nurture their well-being, fitting into the ecosystem of Oasis West and Oasis East (which are managed by Duke Wellness). However, Oasis Perkins is located, of course, right inside of Perkins Library – and its doors don’t close.

You’ll also find occasional events hosted in Oasis Perkins, from Koru Meditation classes to “Tea-laxation” events. Check out the Oasis Perkins webpage to stay up to date on events, or be in touch if your organization would like to host a relevant get together in this space!

You can find a smaller space on the second floor at the Prayer and Meditation Room in Perkins 220.

Other Wellness Resources at Duke

For other tips and events to help you end the semester strong, check out the Duke Libraries End of Semester Survival Guide. There are also a number of resources right on Duke’s campus to support your mental health, which include:

Oasis Perkins has existed in its current form for only one short semester! Is there something that could change about the Oasis Perkins that would help you re-charge? Our team at the Libraries would love to make it better for you. Fill out a feedback from in the suggestion box in Oasis Perkins, or reach out to brenda.yang@duke.edu with your comments or suggestions.

Textbooks and Bean Bags: The 2018 Student Library Satisfaction Survey

This spring, Duke University Libraries conducted the 2018 biennial user satisfaction survey, a large survey of students and faculty at Duke. The goal of the survey is to gauge overall user satisfaction and to gather specific ideas for improvements to DUL materials, services, and spaces. In this post, we’ll share some of the trends within the student responses.

Survey methodology

Since 2013, DUL has created custom surveys rather than use generic survey products, allowing us to customize questions to different patron groups and even different parts of the campus libraries system. Developing and analyzing the results of a customized survey, however, is no small feat! The survey is run every two years, in part because the full cycle of survey development, dissemination, analysis, and follow-up takes the entire two years.

The 2018 survey was deployed in January 2018. A sample of students and faculty received personal invitations over email, but the survey was also advertised on the DUL website and open to anyone. We received responses from 2,610 students. We don’t have full demographic information for everyone, but approximately 54% of the students for whom we have demographics were undergraduates. The survey took approximately five to seven minutes to complete.

Two pie charts. One pie chart shows the distribution of the total 2,610 student participants (48% undergraduate, 41% graduate, 12% unknown). The second pie chart shows the distribution of the 2,307 participants that don't include unknown (54% undergraduate, 46% graduate).
After the survey closed, a group of seven staff at DUL divided up approximately 3,600 free-text responses and manually coded them for topic and, where appropriate, whether they were a request for a new service or change in existing policy or a compliment. The survey data have been visualized in a series of public dashboards. To gather additional information about some of the results, the Assessment & User Experience department also hosted several follow-up focus groups with both students and faculty. The focus group results, while not incorporated into the survey dashboards, have been incorporated into summary reports and recommendations.

The good

“I think the library is one of the places of greatest mutual respect on campus. There is less social stratification and freer flow of interaction. I enjoy my time in the library quite a lot.”

The survey included questions that everyone answered and questions that were specific to different libraries. All survey participants identified which library they visited most frequently. For students, 77% selected Perkins & Bostock as their primary libraries. Only 3% (76 students) reported that they don’t physically visit a library.

A bubble chart showing the libraries visited most frequently by students. Perkins & Bostock Libraries are highest with 77%, followed by Lilly Library with 11% and Divinity Library with 6%.

The libraries are considered an important part of the Duke experience by over 80% of participants. Focusing on the students who picked Perkins & Bostock as their primary libraries, we can look at usage of and satisfaction with the library. Of the 1,978 students who responded, over 80% visit Perkins & Bostock at least once a week. And by and large, students are quite satisfied with Perkins & Bostock. Less than 1% of responses fall in the “not satisfied at all” or “not very satisfied” categories, and the vast majority are very satisfied.

Three related bar charts. The first bar chart shows responses to a question asking students to agree that the library is an important part of their experience. 30% of students selected "somewhat agree," and 51% selected "strongly agree." A second chart shows that, for the Perkins & Bostock Libraries, 18% of students visit once a week, 38% visit more than once a week, and 25% visit daily. The third chart shows that for overall satisfaction with Perkins & Bostock, 14% are somewhat satisfied, 65% are very satisfied, and 21% are extremely satisfied.

The Duke University Libraries value diversity of thought, perspective, experience, and background and are actively committed to a culture of inclusion and respect. Beyond gauging user satisfaction, this year we also asked students about their impressions of Duke and DUL as safe spaces. (In the survey, “safe space” was defined as “a place in which people can feel safe from discrimination, harassment, and any other emotional or physical harm.”) We were excited to find that overall students agree that DUL is a safe space (92% respond with “agree” or “strongly agree”), even more than they agree that Duke University as a whole is a safe space (78% response with “agree” or “strongly agree”). Similarly, when asked if the library is a welcoming place, almost 90% agreed. Despite these encouraging numbers, we are committed to continuing to improve in this area wherever we can.

Three related bar charts. The first chart shows that when asked if Duke University is a safe space, 44% somewhat agree and 34% strongly agree. The second chart shows that when asked if the Duke Libraries are a safe space, 32% somewhat agree and 60% strongly agree. The third chart shows that when asked if the library is a welcoming place, 35% somewhat agree and 53% strongly agree.

The So-So

“I use the libraries a lot to study (esp Bostock) with friends, which is both helpful for me academically and comforting for me socially. The libraries fills up pretty often during busy times, so I wonder if more chairs would help accommodate more students (not even more tables, just more seating). Thanks!”

Even though by-and-large students are satisfied with the libraries, they were not afraid to let us know what areas could be improved! They gave us their constructive criticism in a few ways. First, we asked students to offer their opinions on the possibility of expanding different types of library services. Next, we asked how important specific services, materials, and spaces were, as well as how they were meeting the students’ needs. Finally, we gave them the opportunity to offer additional comments about DUL and suggestions on how to make DUL more of a safe space.

When we asked students what services should be expanded, students were most likely to vote for more spaces for individual study, more spaces for collaborative study, and more textbooks to check out. A second tier of requests include better signage, delivery of items between campuses, lockers, and help with digital scholarship.

A stacked bar chart showing the results to a question about the desirability of various specific services. The most desirable service is "more spaces for quiet or individual study," which 88% of responding students said would improve their experience either a little or a lot. Second is "more spaces for collaborative study" with 80% of responding students. Third is "more textbooks to check out for my classes" with 71% of responding students.

Looking at library-specific responses, we can find a bit more detail about these requests. When looking for services that are both important and not meeting students’ needs, we can see that reservable project/study rooms, a variety of seating options,  adequate quiet study space, and textbooks on reserve all appear in the high quadrant for both importance and not meeting students’ needs.

A scatterplot showing that four services are high on importance and on the percentage of students whose needs are not met: adequate quiet study space; variety of seating options; reservable study/project rooms; and print book, textbooks and articles on reserve for classes.

While not every student followed up on these questions with free-text explanations, the analysis of the free-text comments are consistent with these results. Of the 769 student comments that included requests for new services or a change in policy (rather than compliments), the top code was study/research space, which accounts for approximately 12% of the total requests. The second most frequent code was noise (about 9.5% of the requests), clarifying some of the complaints about “adequate quiet study space.” Requests often include a desire for the Libraries’ quiet space policies to be better enforced. The third most frequent code was atmosphere/sense of welcome – e.g., how inviting the library feels, feelings of “stress in the air.” This code was applied to just over 8% of the requests.

A bar chart showing the top 10 topical tags for requests, including study/research space (about 12%), noise (about 9.5%), and atmosphere/sense of welcome (just over 8%)

Security, furniture, advertising, and signage also ranked highly among requests. Students seem especially desirous of “comfortable” seating; write-in comments mention several types of comfortable seating by name, including couches and bean bags.

The Unknown

“Having taken this survey, I have realized that there are many things which the Duke University Libraries offer which I am not currently taking advantage of…”

While student needs and reactions change over time, one thing remains the same: they unknowingly request services we already offer. Sometimes the survey itself alerts students to particular services.

When we ask students how certain services are meeting their needs or which services should be expanded, we offer a choice labelled “I didn’t know the the library provided this.” Here are some of our most pressing “marketing opportunities,” according to the number of people who were unaware of the service.

A dumbbell plot showing the percentage of students who didn't know about a service, split into those who did find the service important to their research, coursework or teaching and those who did not. Regardless of importance, the service with the highest percentage of students who did not know about it was "support for using, analyzing, and visualizing data." Second highest was "self-checkout stations," followed by "scheduled assistance from library staff."

For each service, there are two values – one for the students who marked the service as important and another for those who didn’t. As might be expected, awareness is always lower among students who don’t find the service important, but there are also services that have lower awareness overall. Services like support analyzing data, self-checkout stations, meetings with library staff, and reservable interview rooms may be good candidates for increased marketing. (If you look at the previous scatterplot, you’ll see that reservable interview rooms also had a high value for students whose needs weren’t being met, even though it’s not rated very highly on importance.)

Another good indicator of marketing opportunities is our analysis of the students’ free-text comments. Some of the major requests from students actually match up well with some of our existing but possibly under-advertised services.

We already know that students are always on the lookout for quiet study spaces. This need is especially pronounced for graduate students, who seem to feel outnumbered by undergraduates, who need quiet space for long periods to work on independent research projects, and who don’t always have private office space elsewhere on campus. When we asked students about services they would like us to expand, we offered them the opportunity to comment on “Additional specialized spaces for honors researchers, graduate students, or other student populations.” Out of 281 total comments on additional specialized spaces, 142 (or almost 51%) mentioned graduate students. In analyzing the comments and in follow-up focus groups with graduate students, however, it appears that many are not aware of either one or both of the dedicated graduate student spaces in Perkins Library.

Riess Graduate Student Reading Room

a photo from outside a room, showing 2 large tables with 8 chairs each and outlets along the top. The room has windows along the back wall and a keypad on the door.

The graduate reading room is a shared reading space for graduate students on the 2nd floor of Perkins. It has a key pad entry code that can be obtained from the Library Service Desk. The room is has good natural lighting and is an “absolutely quiet” zone. Some of the requests indicate that students would like more individual desks, however, so some students may be unsatisfied with this as the only dedicated space open to all graduate students.

Graduate Research Commons

a photo of a room with many cubicles and lockers and a few windows in the back

In the spring of 2016, a large room on the second floor of Perkins was converted into the Graduate Research Commons. The space has 27 individual cubicles of two different heights, adjustable sit-stand desks, and dedicated lockers for all users. The room also includes a technology center with an e-Print terminal, a scanner, and a desktop computer with the Adobe software suite.

Unlike the Graduate Reading Room, however, students must apply for access to the Graduate Research Commons. Despite its many features, the space has been underutilized, and it appears that many students are not familiar with it and have never tried to apply for access. A review of this space could reveal ways to market and set policies for the space.

Next Steps

To determine the most needed and feasible improvements for follow-up, the Assessment & User Experience department will host a DUL-wide staff workshop in July to review the results and make specific recommendations to improve the experience of all of our users. Contact us if you would like information more about this workshop.

We look forward to sharing more of our progress on this and other assessment projects for DUL in the future!