DISAGGREGATED VISUALIZATIONS ~ BFONGDATA

Disaggregated visualizations show data at a lower level of detail. As an example: viewing the sum of sales for all orders versus the sales for each individual order.

Disaggregated visualizations are one of my favorite ways to view data! I don't see these visualizations as often as I would expect. Maybe this is because we are not fully aware of the benefits of them or it could be because tableau likes to aggregate everything.

Some of the benefits of using disaggregated charts:

You can see more data points because you're viewing the data on a more granular level which makes it easier to find outliers, distributions (skew), and relationships.

Before you start using disaggregated data

There can be a lot of data which can affect Tableau’s performance or become messy. As a solution for this, you may need to break up the data by adding filters or use labels/annotations so that users can quickly and easily understand the visualization.

Example 1: Scatterplot

This scatterplot is a great example of a disaggregated viz because the viewer can easily see all product categories at once and make comparisons between them. Adding the annotation (quadrant label) and colors to the quadrants on the right side is helpful because the viewer doesn't have to take time to interpret which quadrants are successful or not.

Example 2: Scatterplot Over Time

This scatterplot visualizes tweets over time and how many favorites they got. Using the tooltip or actions you can show the actual tweets text and links. Using the tooltip you can add any other important information about a single tweet.

Example 3: Unit chart

This unit chart is great because it answers the overarching question, "How many times was a procedure performed in the past 12 months" but it also provides more details. Information on patient risk level is show by color and man other data points for a particular procedure can be added to the tooltip.

Note: used a reference line to add the total bar numbers.

Example 4: Box and Whisker Plot

Box and whisker plots could also be a good choice because you can see the distribution of the data, averages, and easily spot outliers. All of this information is great if you know how to read it, but for some users box and whisker plots can be overwhelming. Make sure your visualization is appropriate for your audience!

A common issue with disaggregated visualizations that you can see in the above graphs is that the points (circles) overlap. To get around the overlapping, I created a second box and whisker that shows a particular occupational group but jitters (randomly moves dots to avoid over crowding of points) the dots (using Steve Wexler's technique) so you can view the points more easily.

As I send you on your way to make your disaggregated masterpieces, my words of caution are, "Just because you can doesn't mean you should!". When creating disaggregated visualization it's easy to get carried away with the marks card. Don't overwhelm your audience! Keep visualizations simple!