Peak Usability in partnership with TERN and QCIF

Please tell us what you did to bring order to big data and what effects your solution had, and help us understand what makes this solution noteworthy or exceptional.: 

The project
Usability testing, user experience research, collaborative design of the TERN Data Discovery Portal

The client
The Terrestrial Ecosystem Research Network (TERN) connects ecosystem scientists and enables them to collect, contribute, store, share and integrate scientific data from multiple facilities and databases across Australia. Collectively this increases the capacity of the Australian ecosystem science community to advance science and contribute to effective management and sustainable use of our ecosystems.

The brief
TERN were about to launch a new data discovery portal and wanted to make sure it was addressing the needs of its users. They wanted to understand what users wanted or needed from a scientific data research portal, and how the data would be used and accessed. They also needed a design that is visually appealing, inspires confidence and meets users' needs. The scientific data contained in the TERN DDP is massive and complex and includes: spatial and temporal data, species data, vegetation data, soil data, thematic data, ecological, atmospheric, and geographical data, to name a few.

What we did

  • Contextual user interviews with ecologists to gain an in-depth understanding of data needs, motivators, environment and existing issues experienced by academic and scientific users.
  • Collaborative design workshops with academic and research staff, Government stakeholders and members from the scientific research community.
  • One on one usability testing of the beta version of the site with representative users.
  • Wireframe design * Final quality visual design and front end development of design in CSS and HTML5. **The outcome** The launch of the TERN data discovery portal was a success, they were able to provide scientists and researchers with a search portal that is easy to use. We based our designs on the user research and have helped provide a tool to the scientific community that is actually useful and usable, while at the same time bringing order to big data.
What can others learn from the successes and failures of the way you’ve unlocked the value of big data?: 

The key factor that underpinned the success of the design was commencing the process with good user research to understand how users find and use data. Through employing contextual interviews and collaborative design workshops with users, we discovered:

  • what search parameters they used to search data and to what level of granularity;
  • what specific spatial search parameters were important;
  • what issues and frustrations they had currently locating and accessing data;
  • what metadata they wanted to see;
  • what format they wanted to access data e.g. csv, database, MySQL etc; and
  • how they used data e.g. data modelling climate. 

While this initial step is not unique in itself in the user experience field, the outcomes were fundamentally critical in shaping access to the data. We also asked users what other websites they used outside of their field and if there were any design conventions they felt could be applied to this data discovery portal. What we found is that every user is different and wants to search for data using different search parameters. Therefore, a single advanced search or drill down browsing approach was not going to work. Instead, we used the faceted search approach common to many ecommerce websites such as eBay which allow users to turn search criteria on and off in a side bar and to also see the number of results for each search parameter. This approach worked really well to overcome the challenges of diversity in ordering big data and accommodate different users with different needs iand the cross-discipline nature of the data and users.

The user research was also able to shed light on the common and high priority facets that were meaningful across users and disciplines. This fed into such design decisions as the facets used, and what facets should be expanded by default. Without this user insight, the data ‘discoverability’ would have been considerably limited.

What were your expectations of the value hidden in your data, and how did they influence the design of your solution?: 

The Terrestrial Ecosystem Research Network (TERN) aims to connects ecosystem scientists and enables them to collect, contribute, store, share and integrate data across disciplines. Collectively this increases the capacity of the Australian ecosystem science community to advance science and contribute to effective management and sustainable use of our ecosystems.

TERN had the foresight to know that the value hidden in the data was largely limited by the discoverability of datasets by the scientific community. The idea of the TERN Data Discovery Portal was to delivering open access to Australia's terrestrial ecosystem data and to make it easier for the scientific community to discover and share data. In order to ‘connect the dots’ and make the data ‘discoverable’, the ‘dots’ that were relevant to scientists had to be uncovered. This is why from the beginning it was clear that unlocking the user insights in the initial user research phase was pivotal to the design. As a result it was imperative to ensure a large degree of user input within the constraints of budget and time.

Accordingly, our approached utilized highly qualitative methods (contextual inquiries, workshops, usability testing) that provided a large amount of anecdotal, ‘in their own words’ data that could be sifted and sorted to pull out the key themes and needs. The design was influenced by this initial focus on understanding the user’s needs.

Conversely, how did the design of your solution affect your understanding of the potential value of your data?: 

After the initial user research, collaborative design workshops were conducted with academic and research staff, government stakeholders and members from the scientific research community. This was built into the design process in order to share the user insights with key stakeholders, scope out design challenges and constraints within the broader business environment and stimulate discussion around various design solutions.

While the user insights remained stable, the solutions generated were varied, many of which had not been considered prior to the session and without the cross-pollination of ideas between stakeholders. For example, during a collaborative design workshop a shopping cart idea was proposed by one scientist for shortlisting data sets. This analogy was well understood by other participants in the workshop and generated positive feedback when fed back to stakeholders and other users during this process. In time this fed into the design of the system that would allow users to populate a shopping cart like widget and “Add to favorites” datasets they were interested in.

Describe the aspects of the design of your solution that do the most to expose meaning in data that would otherwise be harder to discern.: 

Historically, scientists and researchers discover relevant datasets from other organisations and researchers through their networks and at conferences as well as referring to citations in journal articles. This is a very ad hoc approach to locating data and exposing the meaning and value of different datasets.

The TERN data discovery portal’s aim is to help the scientific community discover data and publish high quality metadata to help scientists locate reliable datasets for data modelling and other research without the need to rely on their networks and word of mouth. However, it was critical the TERN Data Discovery Portal is easy to use and allowed users to easily locate data based on parameters that are relevant to them. In the collaborative design workshops, visual and spatial design solutions to give meaning to data were favored by all stakeholders.

Maps, particularly a map of Australia, were proposed by most participants as a means to explore ecological datasets. Utilising a map based search interface and feedback gave a strong frame of reference that was common to all users. This became the basis for the design solution. However, spatial parameters were only one facet. Users were also interested in applying other parameters that allowed them to locate all scientific data collected at a particular geographic location at any single point in time. Hence, in theory, scientists could locate soil data, vegetation data, fauna and atmospheric data for a particular latitude/longitude at a specific point in time. Discoverability of these datasets will hopefully provide opportunities for the scientific community to conduct more accurate data modelling of ecological data which has a number of potential applications.

How might your solution be extended or adapted to address additional types of data and other questions?: 

It is envisaged this solution can be expanded by incorporating more data sets generally, as well as from other scientific disciplines. There is also potential for data modelling capabilities which could attempt to provide more advanced statistical analysis of the data. TERN are also in the process of developing a portal that not only allows users to discover data but also to access, query extract data in certain formats.