This project was completed as part of coursework for my MSc in Human-Computer Interaction. We designed a campus navigation app, with a user-centered approach.
I collaborated with Kate Johnson, Iris Lau, and Bowei Zhang on this project.
We were tasked with designing a navigation app for the University of York campus, with a focus on 3 key features that emerged from research. To do this, we used a range of methods including:
- In-person interviews,
- Scenario based design
- Heuristic evaluation
- Interviews with paper prototypes
Our very first step was to conduct a series of 1 on 1 interviews with potential users of the campus navigation app. We interviewed 11 students and 1 staff member.
We used semi-structured interviews. This means we had a set of questions to follow, but we could deviate from that structure and ask additional questions when appropriate. Semi-structured interviews are great at teasing out the ‘hows’ and ‘whys’ of how people use applications, because you can ask lots of follow up questions.
Our interviews revolved around 2 main topics:
- How they currently use navigation apps.
For example we asked them to show us how they would navigate to a place, what they didn’t like about the experience.
- How they currently found locations on campus.
For example, we asked about previous experiences trying to find rooms in unfamiliar buildings on campus.
We then used affinity mapping to group and prioritize our results.
We learned a lot about what mattered to navigating on campus:
- Finding the quickest route between two locations
- Detailed, mapped information of rooms and shortcuts
- Information about which paths are well illuminated at night
- Better GPS service on campus
- Wide range of features such as offline maps, mid-route changes etc.
We then focused on the most common needs and used those to build personas. The personas were designed to be diverse in many ways, such as gender, age, ethnicity, occupation, digital inclusion etc.
Here is an example of one of the personas we created:
Scenario based design
Next we used scenario based design to understand how someone like Yi currently navigates on campus, and what additional features could help her.
- Current scenarios based on multiple personas
- Claims analyses of these scenarios ( specific positives/ negatives)
- To-be scenarios, with possible solutions to current negatives
Based on our user interviews and persona scenarios, we came up with the following features to test a prototype:
- Search for a location within a building.
- Provide directions between the user’s specified location and a location of their choice.
- Allow the user to tailor their route on specific criteria, including transport method and how well-lit a route is.
- Help users find entrances to large university buildings.
We started out with pen and paper wireframes. Here are some that we made:
Then, we made a higher resolution prototype using Figma.
We evaluated the prototype using Petrie and Power’s usability heuristics. This is basically a process of identifying usability issues, and rating how important those usability issues are. When we had reached a consensus on the most important issues, we corrected those problems in the designs.
Finally, we brought the prototype to a new set of participants. We opted to use paper prototypes at this stage so we could focus on the features and not on the limitations imposed by most prototyping tools (e.g. no ‘pinch-to-zoom’ on maps).
We did in-person usability testing. We had a series of tasks for participants, and would simulate screens changing with the paper prototype. We asked participants to ‘speak their thoughts aloud’, per the ‘think-aloud’ protocol.
Participants highlighted usability issues, and we asked participants to also rate how severe those issues were. Based on these ratings, we made several changes to the designs and functionality.
For example, one of the desired features was an option to search for well-lit paths at night, as some areas of the campus are dark and isolated. In our usability testing with participants, several participants highlighted that that they wanted better feedback to know when the feature was active:
Our final design included a range of features that were suggested and validated by participants. A few of the screens are below
Click the images for higher resolution.