In this project, my team was tasked with conducint a usability evaluation of the Purdue Libraries website and suggesting changes to improve the experience of undergraduate users.
Role: UX researcher, Project lead
Timeline: Aug-Dec 2023
Class: Experience Studio 2
In experience studios we work on one semester long project with a corporate sponsor and a group of UX students from every year of the program. Students in experience studio 2 tend to be sophomores.
Purdue's Library has access to both physical and digital materials, and both of these types can be searched through the Purdue Library's website. The website is run using a database service called PrimoVE. This service helps many different libraries across the world organize and access library materials.
In this project, Purdue Libraries tasked my team with evaluating the current website for usability, and making recommendations for changes that would comply with PrimoVE's limitations.
Heuristic issues discovered with layout, terminology, and lack of visual indicators
PrimoVE only allows for certain formats, but libraries do have some flexibility in their sites design
Search facets and advanced search were found to be confusing or worked around by participants
With the help of our project sponsors, Purdue Library representatives, and an impact-effort matrix, we narrowed our scope to focus on improving search facets.
After deciding to focus on improving search facets, we conducted a series of interviews to determine user opinions. These interviews included terminology testing, card sorting to determine facet importance, and an aesthetic comparison to learn what PrimoVE formats users preferred.
Redundancy and confusing terminology were the biggest problems with facet titles
Participants preferred the style of Vanderbilt's PrimoVE site because its facets were shown in boxes
Some Facets could be deleted or combined
Aesthetic Changes
- Facets were given a background based on the aesthetic comparisons from interview comparisons
Facets Removed
- author, citation source, external data source, and external search removed as suggested by interview participants
Facets Condensed
- Facets with similar purposes as defined in our interviews were placed under the same drop down menu
Facet Titles
- Facet titles that took up more than one line were condensed for visual similarity
Purdue's Library has access to both physical and digital materials, and both of these types can be searched through the Purdue Library's website. The website is run using a database service called PrimoVE. This service helps many different libraries across the world organize and access library materials.
In this project, Purdue Libraries tasked my team with evaluating the current website for usability, and making recommendations for changes that would comply with PrimoVE's limitations.
To guide our project, my team created 3 goals with guiding questions that would help us create a successful end product. These goals would be used to split up our activities and achievements to keep our project on the right path.
1. Discover how Purdue Library's Website is Currently Used
a. How usable and intuitive is the website?
b. How do students currently use the site?
2. Learn What Features Users Expect to See
a. What features do students want to see from this type of resource?
b. What changes are feasible to implement?
3. Design for Identified Opportunities
a. How can we reduce frustration and increase ease of use for a database search website?
b. How can we alter the current system while staying within system limits?
The brief of this project was very broad. Purdue Libraries asked us to improve the user experience of the website which meant there were many directions we could go.
Our first step in the design process was to discover what pain points existed in the current service, using a variety of tools, we were able to pick out objective and user-sourced pain points.
A heuristic evaluation is a relatively quick and objective way to pick out usability issues.
Using the 10 heuristics from Neilson and Norman, our team discovered issues with functionality, aesthetics, and terminology.
We found many small violations covering most of the 10 heuristics, but below are some of the most noticeable to our team.
Few icons used, many sections only identified with text reduces scannability - Recognition rather than recall
Website is not responsive to changing screen sizes, creates confusing visuals - Flexibility and efficiency of use
Confusing terminology and jargon used frequently throughout the site - Match between systems and real world
Differences in search and advanced search that aren't explained - Consistency and standards
Purdue Library's website exists as a way to search and browse Purdue's resource database. To learn how the features of the website compared to other schools who used PrimoVE, as well as other resource database searches, we conducted a competitive analysis.
Facet (search modifier) hierarchy and quantity varies greatly between sources.
Use of specialized terms or general terms varied between sites. General terms tend to be more understandable.
PrimoVE sites follow a similar layout with similar features, but some websites utilize those features better.
Seeing all of these differences between sites, even between those that use PrimoVE, we decided we needed to test how these features impacted the user experience of both experienced and inexperienced students.
Purpose: Discover which features and parts of the Purdue Library website create the most pain points
Participants: 4 Experienced users (have used the website 5+ times) & 2 unexperienced users (no prior use)
Method: Participants were asked to go through a series of tasks that would be part of a normal search process. Participants were encouraged to think aloud and their words and actions were recorded and analyzed.
Many participants found the advanced search confusing. Participants tended to stop using advanced search half way through or did not use it at all.
Most participants struggled to use the side bar containing search facets to complete complex tasks.
Many participants expressed that the wording was too long in many places
At this point in the project we had discovered many pain points and it was necessary to narrow our possible design spaces down so we could select a space to design in.
In this workshop we worked with our project sponsors, two representatives from Purdue Libraries, to determine what design spaces were possible. To do this we used an impact-effort matrix.
Several solutions were placed in the Quick Wins sections. Small tasks like editing the ask us button and icon usage could be easily changed, but larger issues like the facet organization would take more effort with a big payoff.
We decided to spend the rest of our time working to do the research to address the facet hierarchy while creating recommendations for the smaller issues.
Facet organization was the most solvable issue we discovered that still needed research. To do this we started off with secondary research, a more focused competitive analysis, and a heuristic evaluation.
Findings:
- Simplicity leads to more decision making and engagement from the user
- Users feel more risk when presented with more options
- Especially in a search context, users might feel that tinkering and using the facets could waste time
Findings:
- Terminology was not succinct or clearly defined
To test terminology, order, and look, we had 8 participants complete a series of activities.
Participants: 6 Participants were experienced with the website (used it 5+ times), and 2 were not (have not used it)
Terminology Testing
Participants were asked to describe what they thought each facet would do based on the name
Card Sorting
Participants were asked to sort the facets in order of importance to them in their content searches
Aesthetic/Layout Testing
Participants looked at three different PrimoVE supported websites that had different layouts and aesthetics. The Purdue, Harvard, and Vanderbilt websites were used.
Harvard Library Website
Vanderbilt Library Website
Purdue Library Website
1. Publication date was important to 2 of our users but couldn’t find it on the site
2. Vanderbilt was considered the simplest
3. Redundancy and vagueness are the main issues
4. Terminology for some of the facets can be simplified
5. People liked how the “show only” is open by default
6. Main subjects used were date of publication, key words, and type of source (online vs not)Material type is very important to people
Aesthetic Changes
- Facets were given a background based on the aesthetic comparisons from interview comparisons
Facets Removed
- author, citation source, external data source, and external search removed as suggested by interview participants
Facets Condensed
- Facets with similar purposes as defined in our interviews were placed under the same drop down menu
Facet Titles
- Facet titles that took up more than one line were condensed for visual similarity