Government Open Data Interaction Design Guide
Making data simple for citizens.
Government open data initiatives are still in it’s early stages of maturity. Cities and local government agencies are making steady progress on the publication of quality cleaned data on open data portals.
As publication methods evolve, focus and innovation is now needed on awareness of government run activities for the very citizens that generate and fund government open data systems.
The following is a data interaction guide that incorporates human centred design to create meaningful data interaction with citizens that might not understand or be able to read the data.
The motivation of this guide is for government bodies, the open data community, hackers and designers. To build products to increase education and awareness of how our governments are run and bring data appropriately to people in the right format.
The data interaction guide is for any individuals or agencies interested in creating citizen centred data driven solutions. The data interaction process can also be used when designing user experiences for non technical users interacting with data visualiziaons of interactions.
The guide is intended for interaction designers and product managers with some knowledge in the UX design process.
Data Interaction Design Guide
The following guide will help organise the design process when a data driven solution is the appropriate choice to solve a problem. The Guide is broken down into 4 Steps. Each section also has information on how to use human centered design principles to create data interaction suited for the specific audience.
5. Test and Iterate
Start with Cleaned Data
The process starts with clean machine readable open government raw data.
General software knowledge in Excel and Tableau (Tableau Public is a free service but the date is not private) is advised to read and manipulate input data into visualisations for testing purposes.
Human Centered Research
Having user empathy will create a more memorable user experience. Understanding the technical capabilities as well as the data literacy of the user will determine the complexity of the data interaction and communication designed.
Define Data Story
Input data into Excel to view how the information will need to viewed by the user. Data can be interpreted in various ways. Decide which data can remain static vs interactive by the user. Evaluate the story the data is presenting to the user.
Play with the data
Input the data into a visualisation software such as Tableau to test various methods in interaction and presentation.
Define what the end user data tasks will be. For example do they need static data or interactive data? How prior information do they need to understand the data interaction?
Human Centered Interaction
This is also the time to research the user and understand what their needs are. Do they need access to all the data or a smaller section? User needs should always be the focus.
Visualise with Real Data
Actual data is advised to use testing interactions to reduce issues once the real data api is implemented.
Sketching interactions helps test the various methods of telling the data story. Test this process to make sure the interaction is functional.
Human Centered Interaction
Data is non communicable for most people it is vital that data is translated in a way that is understood by users. Users might not know how to read financial data and will need help in translating the information to suit their level of knowledge.
Edit & Curate Data
Too much information can cause confusion and misunderstanding. What data is relevant of the task of the user? Editing data to fit the user need will help is reducing cognitive load.
Wireframing the interaction provides a tangible flow of the design to help communicate the process of the interaction. This can be done on Sketch or Illustrator. Lo fidelity paper sketches on can also be tested
Prototyping helps in assessing what is working and what is not with the interaction. This can then be tested on users to make sure the interaction of the data is functioning as it is intended to. Invision is a great tool for easy interaction testing.
Human Centered Communication
Creating an easy to use guide on how to read the data might seem redundant but users will benefit from a level of education on how the government works and what it is providing to give them a level of confidence in data legibility.
Visual design is an important component to the process. It is imperative that once the overall User experience concept is completed, visual design relevant to the user is implemented. Remember not to make this pretty or beautiful, it should be function and then aesthetically pleasing. Minimal viable products will be helpful if the data is only to be used for a smaller number of people.
Language Is as important as visual design. The language needs to reflect the user and how they will best understand the data communicated to them. Choosing words like select and compare might seem simple but if the user does not know what to compare or what the data means it will be useful. Using correct language for the user will make their life easier and the interaction more effective.
Testing and Iteration
Design with Iterations
Design with small iterations. Designs will need to be iterated so work on the core of the design and iterate as new features are embedded.
Test at each stage
User testing can be conducted informally with any user in the initial stage. The purpose is to test often to make sure the design is functional for others. Testing should be conducted with actual users before the presentation phase.
Second Product Iteration
After delivering the final product, second product iterations and user testing can add new features and build on the minimal viable product delivered.
This is part of a research project conducted for a masters of Experience Design candidacy at Hyper Island UK. The following design guide is based on a combination of expert interviews in both the UK and the U.S open government data community with personal experience in open government data interaction.
I plan to develop this guide further and break down the components in further posts.
Follow me on twitter to keep updates @sarkyou
Feel free to leave me comments and feedback so this process guide can be improved for the benefit of all working with data.