My #BestNewTableauSkill this week focuses on a technique I learnt through people's #MakeoverMonday and #RealTableau submissions; not to mention it being used in the Big Book of Dashboards; creating Jitter Plots.


A Jitter plot is similar to a box plot however it displays the data points across a random y axis; allowing you to easily distinguish multiple points in a way that would be impossible if they were plotted on a single line.


To create a jitter plot is surprisingly simple. First create a standard box plot by placing your measure on Columns and your dimension in the details marks card. Next create a new calculated field random() and place this calculated field on Rows; it doesn't matter what aggregation you select but for simplicity I have placed it as a Continuous Dimension.


if you wish to place a box plot over the visualisation you can do using the analytics tab but this is purely a matter of personal choice; in the example below I have used one.


It is as simple as that; just one extra step and your box plots can become a lot simpler to read when you are visualising a large number of data points; the examples below hopefully evidence the difference.


Traditional box plot

Traditional box.png


Jitter plot



Simon (@SimonBeaumont04)