Our understanding of Alteryx, Tableau Server, and Tableau was put to the test on the first day. The information we were given was about the NYC Council Constituent Services. Our goal for the day was to reproduce the charts and address the difficulties with accessibility. More information on the dataset may be found here.

Step 1 Alteryx

It was necessary to use the API to get all of the data from the website. We were able to retrieve all of the data from the website by using Alteryx's download data tool. The following step was to parse the data in order to gather it in a clean, readable format (equivalent to the one on the website). From Text to Columns, Datetime to Crosstab, and Data Cleansing, Alteryx tools were used to parse and clean the data (alternatively you can use JSON Parse to parse the data). The workflow is shown in the image below.

This is how the final table looks like:

After obtaining the desired table, the following step was to publish the dataset to Tableau Server. The new feature I discovered today, aptly titled "Publish to Tableau Server," allows you to publish data directly to a Tableau Server in the Hyper Extract format (You never fail to amaze me, Alteryx!).

Step 2 Tableau

It was now time to use Tableau to create the charts and make them more accessible to a wider audience. Here, I chose to create two dashboards, one for typical users to navigate with a mouse and the other for people who only have access to a keyboard. We also had to make sure that the screen reader read things in the correct order.

The following are some of the techniques I used to make the dashboard more accessible:

1. For each chart, include dynamic captions.

2. Create distinct dashboards for different users.

3. Include commands in the subtitles to access the underlying data.

4. To improve visibility, add markers to the line chart.

5. Instead of using the canvas, add text to titles so that the screen reader can read it.

6. Use a colour palette that is appropriate for a wide range of audiences.

7. Add gridlines to make it easy for users to reference the values on the axis.

8. Make use of dynamic BANs

9. To distinguish similar-looking characters, use the 'Verdana' font (Ex I and 1).

10. Finally, instead of drill downs, show filters as a dropdown as keyboard-only users cannot interact with charts directly.

It was time to upload the dashboards to the server to check the focus order (the order in which the screen reader narrates the data) and test, optimise, and reiterate the process until we were satisfied with the focus order.

The Dashboards can be found here.

THE MAIN TAKEAWAY

A dashboard is valuable when it reaches a larger audience, but it is most efficient when it reaches the entire audience.