As part of the project to improve MyEd, we’ve been focusing on how content is organised. We worked with the UX team to use card sorting studies to help us research how our student users understand and group content. We used our findings to re-organise how content is described and presented on MyEd, and develop the first iteration of a new information architecture and navigation structure.

Why card sorting?

We already knew a lot about the content on MyEd, and which tasks are most important to students. We also knew that the current structure wasn’t working.

Card sorting is a simple way to find out how students would group and structure content, and also how well they understand the content and how it is described.

The UX team designed a series of card sorting workshops and studies which would allow us to get as much information as possible within a restricted time frame.

Initial workshop

The first task was to reach a starting point which we could use for further research. We started by inviting a group of students to an open sort session. We had sets of 70 cards describing the content in MyEd, with each card describing one task or piece of information in neutral descriptive language.

The students were asked to work in pairs, and each pair to group the cards into categories that made sense to them. They were asked to work collaboratively, discussing their thinking with each other as they went, and create groupings that made sense to both of them. They could group the cards any way that they wanted, create new cards, relabel the cards and also discard cards that they didn’t think were important.

Once they had grouped the cards, we also asked them to label the groups.  A member of the project team worked with each pair to provide support, and also ensure that both students participated, and that they collaborated on and discussed their decisions. This also allowed insight into why the students grouped the cards in certain ways.

At the end of the first workshop, we had three sets of grouped cards.  After analysing the groupings, and sharing the discussion points, it was clear that:

• The most important content/tasks were those which students had to do most frequently, or which were important to their studies
• The largest cluster of content was information and resources relating to the students’ own current courses
• All the groupings included at least one level of sub-group below the top-level groups
• All the groupings pulled together information relating to payments into a single finance/money section
• Generally, the students understood the descriptions, and did not relabel many cards.

After the workshop, the project team met to consolidate the results. We produced a new set of cards, including the new cards which students had created, and any revised descriptions. We reviewed the groupings that the three groups created, and standardised them into a single top-level grouping:

• Priority 1 (a neutral way to identify items that belong on the home or start page)
• Studies
• Library
• Money
• Life and Career
• Accommodation
• Account

This provided a starting point for the next phase of research.

Follow-up workshops

For our next phase of research, we wanted to validate and refine the groupings, with the goal of achieving a fairly stable model that we could then validate through a large scale online exercise. In order to do this, we ran three further workshops, each with one pair of students. This time, students were asked to review and modify the existing groupings, and modify them, rather than starting from scratch.

We gave the first pair of students the revised cards and the standardised top-level grouping, and asked them to organise the cards. They could change the top-level grouping, add new groups and sub-groups, create new cards, relabel cards and discard them.

We then took the groupings and card set that the first pair of students created, and asked the students in the next workshop to repeat the exercise using this new grouping and card set, and then did the same for the last workshop. The sessions were moderated, and again, the students were asked to collaborate and discuss their decisions.

The first pair of students made quite significant changes to the groupings. They created three new categories: Your  Courses, EUSA and Community. They combined Money into Account, and they separated Career into its own category.

The second pair of students removed the EUSA and Community categories, instead creating a Student Life category.  They also made minor changes to the labelling of other groups, but didn’t make any more major changes.

The final pair of students combined Library with the category initially labelled as Studies, then relabelled as Learning Tools, and called it Learning Resources.  Again, they made some minor changes to the labelling of other groups, but didn’t make any more major changes. This final revision gave us the following top level groups:

• Front Page
• Courses
• Your Accounts
• Learning Resources
• Student Life
• Careers

Online card sort

The workshops provided a starting point for a new information architecture for MyEd, but this was based on research with a small number of students. The next step was to test the proposal with a large group of students, so we set up an online study, open to all students. Based on our observations in the workshops, we decided to make one change to the top-level groupings, combining Courses and Learning Resources into a single group called Studies, which included the sub-groups from both categories.

The online study was a hybrid card sort. Participants were given the groups and sub-groups that emerged from the final workshop, but were also able to create their own groups and relabel groups. They were given a set of 80 cards, which included the new cards created in the workshops, and which used the revised descriptions that students had created. Participants were asked to sort the cards into the groups that made sense to them.

Over 1000 students participated in the study with 536 students completing it. This proved challenging to analyse, but produced very interesting results. The categories were reviewed, and duplicate and similar categories combined. Cards were then sorted to identify the categories into which they were placed most frequently. This was done for the whole sample, and also across different segments, for example split by level of study (undergraduate, postgraduate taught, postgraduate research).

The highest frequencies were all in the categories that were pre-populated in the study. This was the case overall, and also when the different segments were analysed. However, although the categories for the highest frequencies were very consistent, only 43 of the 80 cards were placed in the same category by 50% or more of the participants. 15 of the cards were placed in the same category by less than 30% of the participants. This meant that we also needed to consider second, and in some cases third choice placements in our analysis.

There was a high level of consistency across the different segments. We had expected that there might be significant variation across different student groups, for example undergraduate and postgraduate students, first year students and students that have been at the University a year or longer, UK students and students from outside the UK. In fact, results were broadly similar. 68 of the 80 cards’ most frequent placements were identical across all segments. 11 of the remaining cards were split across two categories, with only 1 card split over more than two categories.

A draft information architecture and navigation

Using the data from the card sort workshops, and the online study, we produced an initial information architecture and navigation structure for the student view of MyEd.

The next step is to validate this, and confirm that students are able to easily locate information using the new navigation structure. This will be done via an online tree test, which will be written up in a separate post.