For the start of the notorious Dashboard Week, we were tasked with creating an accessibility-focused dashboard. Half of the day was spent talking about what accessibility means in the context of design and how tools like screen readers interact with Tableau. We learned a lot of tips, tricks and considerations to make when creating dashboards that can be effectively used by all sorts of people.

 The highlight of the day was having to try to access a finance dashboard without using the mouse or trackpad. While I was eventually able to view the dashboard and switch between different tabs, it was very difficult to actually glean the information I wanted to see. Going through that exercise, if even just a little bit, gave me some insight as to how challenging exploring dashboards (and using computers, in general) can be without conventional tools like mouses.

 With that said, the rest of the day was spent building the dashboard. For this, we used a dataset that represents pedestrian complaints about streets in Glasgow, Scotland. The data was publicly available and quite sparse. In it, there is latitude and longitude values, the names of different places, the category of street problem (pothole or pavement/footpath related issue) and a datazone field.

 This is the first draft dashboard that I came up with.

 As you can see, there is plenty of room for improvement in the way of making this dashboard more accessible! A few examples are:

1.  Reducing the number of colors used. Right now, 7 unique colors are used which may be visually distracting. Bringing this down would help simplify the dashboard and avoid confusion.
   
   
2.  Labels should be increased so that it is easier to read and interpret information.
   
   
3.  The colors for the category (potholes + pavements/footpaths) clash with the white background and the black text labels. So, it would be ideal to find colors that contrast better with white and black. Alternatively, the white background could be changed as well.
   
   
4.  The map is arguably unnecessary as its core functionality relies on a mouse (being able to zoom in and out). So, in the second draft, this map will be removed and replaced with another graph.
   
   
5.  Like the map, the pie chart is arguably unnecessary. Instead, KPI cards can be used to indicate the total number of street problems (including by category). This would help make the dashboard more simple.
   
   
6.  A parameter can be added so that users can control how many locations they want to see. While 5 will remain the default, it would encourage user engagement if they had the choice to change this number.
   
   
7.  Remove the broken “Download to Excel’ (Download to Crosstab) button. I’ll try to re-incorporate that button if I can get it to work, but for now, it’s distracting.
   
   
8.  In this original design, I included captions but had them “hidden” (white text, font size of 1) — in the next draft, I will show these captions just in case.
   
   
9.  Try using the dashboard without my mouse or trackpad, instead using a screen reader and shortcut keys.

 Applying Universal Design principles to ensure that neurodivergent people are able to interact with dashboards is important. While a lot of the design considerations between accessibility and general business intelligence overlap, greater care needs to be placed on making data products that are truly accessible to all. With my experience today, and further practice, I am looking forward to contributing to the creation and adoption of data products that are more mindful and useful for wider audiences!