Processes

Experiments can go wrong. At the DigiKAR project we tried to embrace this and tried to learn as much as possible from the mistakes we made throughtout the project. With the help of documentation and user tests, we tried get better each time, when creating visualizations for DigiKAR.

How to implement an experimental approach?

For the DigiKAR project we used the following methods:

• documenting (e.g. video-recordings of a workshop, version control)
• iterative process (users, research questions, visualizations)
• sketching
• prototyping

User studies

In our web-based visualisation development, we integrated user studies at earlier stages. With such a user-centred approach we aimed to test new concepts and improve the usability of the visualisation based on the user feedback.

We carried out two different user studies: the first mainly focused on the perceptual and the second on the usability level.

First user study

In the first user study, we evaluated perceptual aspects of developed point-based symbol variants of different visual complexity. The aim was to find out the symbol which performed the best according to the selected criteria (accuracy, certainty, readability, overall performance, aesthetics, and speed). This study had a quantitative character. The sample of 60 participants was diverse, i.e. different age groups, backgrounds, map usage experiences, etc. For this study, we used a semi-structured questionnaire in combination with map-related tasks and user-based symbol ratings.

As a result, the so-called ‘Snowflake’ symbol performed best. Therefore, we continued development of our web-based visualisation using this symbol.

You can find more information on this user study in the blogpost Better safe than sorry – why we are testing visualisations early on the IfL blog.

Second user study

In the second user study (still ongoing), we focus on the usability aspects of our interactive web visualisation and its user interface.

We aim to test implemented interactions, functionalities, and visual aspects and find out the weak points of the visualisation from the user’s perspective.

The second user study is qualitative with an iterative character. Participants (this time with historical knowledge) are asked to solve several map-related tasks, think aloud along the process and in the end, answer open questions related to their user experience.

After each round of the test, we modify and update the visualisation based on the user feedback.