The STEM Transfer Student Success Initiative (t-STEM) is a collaboration between University of Maryland Baltimore County (UMBC) and partner 2-year institutions that was funded by a grant from the Bill and Melinda Gates Foundation. The purpose of the initiative is to aid and increase the success of STEM students transferring from community college to a 4-year institute. 

As the project progressed, more and more content was added as the various subject matter groups collaborating between institutions contributed their work. We eventually went through five different approaches for format, structure, and information architecture as we redesigned for every major addition and shift in content. 

The planned content focused on encouraging students to pursue STEM majors and careers. I created a one page layout and chunked information into panels so that it was easy to read. Later on, when we were given content from the Advising group, this structure no longer worked. The Advising content group felt like they had to cram all of their information into a panel the size of a Powerpoint slide, hiding everything in tabs and multiple columns. Not only did this not look terrible, it was tucking away important information, making it less visible and harder for students to find.

The focus of the website shifted from being an introduction to STEM majors and careers to an online guide for STEM transfer students. Since t-STEM was originally planned as a course to be taught in a physical classroom with print material, the structure and visual style was influenced by what the print material may have looked like. Each section was treated somewhat like a page (with a secondary column that acted like a wide margin for notes), though all the “pages” were connected so that it would still be a one-pager design. The addition of more content made the one-pager layout too long and difficult to skim and scroll for specific information.

I learned my lesson from the ridiculously long one-pagers and broke the content into chunks. We were influenced by GOV.UK which consolidated tons of content and used a drill down navigation pattern to make information easy to find. The concept was that each section would have a main page that would then lead to smaller articles. Unfortunately, some sections had no real hierarchy or would end up having too many levels, so this structure did not work.

After the final layout was approved and all of the sections were established, there was a concern if the information architecture was unintuitive and that the Resources section was too heavy. A content audit was conducted and I explored different approaches to the information architecture.

The approaches included: content organized by the stages a student would go through, from introduction to STEM majors, to transferring, to getting acclimated to 4-year institution; content organized by topics, including transfer process, campus life, academic studies, and career; and guide content separated from resources and organized by topics.

In the Spring 2016 semester, I took the class HCC 729: Human-Centered Design. I saw an opportunity to conduct a usability test for the t-STEM site. Up until then, we had only had informal discussions to get feedback from users, so I was excited to do a formal evaluation on the site, particularly on the organization of the content.

Multiple usability methods were utilized, including: competitive analysis, personas, user analysis, task analysis, environmental analysis, direct observation, interviews, and heuristic evaluations. The results of these methods lead to low-fidelity prototypes that were tested using a think aloud method and then used for participatory design. Medium fidelity prototypes were designed based on the participatory designs to address the identified issues. Usability tests were then conducted to evaluate both the new design prototype as well as the original site.

Although we used draft content for the initial designs, we quickly learned that it was nowhere close to the final content.

Having a full understanding of all of the content before starting the design process would have given us opportunity create the appropriate structure and organization from the beginning, rather than adding content as we received it and having to change the structure when the content no longer fit.

Another issue was that at the beginning, all the content was expected to be static, so we did not expect to have content that would need to be updated every semester by our team. In the end, the constantly changing content was not sustainable and was eventually cut or moved to another departmental site within UMBC. This could have been avoided if we had better planning from the start of the process and included a CMS.

In the future, I plan to make sure to have at least draft content for the entire site before starting the design and determine with the client if a CMS is necessary for maintaining the site.