Improving your Scatterplot - Part I
I these two following blog-posts, I'll try to give you ideas on how to improve your scatter plots.
In this blog-post, I will show you how to improve your scatter plot by utilizing a dual-axis chart as an alternative to the traditional size legend, hereby creating a visual easier readable scatter plot. In the follow-up post, I am going to show how you can color each of your quadrants according to the average of the two measures in your view. This is useful as your can categorize your customers according these two measures. You can then go on and apply different strategies when targeting each specific group.
Alternative Size Legend
The size-legend is one of the more difficult things to use when analyzing data on a scatter plot. The reason is that it draws the attention away from the scatter plot, when we want to add some context to the observations. So to improve the scatter plot we can make use of a dual-axis chart and create an alternative size-legend.
The first step is to create the scatter plot that you are interested and duplicate it:
From here you need to make a table calculation, calculating the window maximum of your second view:
And you need to control the direction such that it computes using the marks in your view - in this case Sub-Category:
Now the trick is to make the two visualizations collapse into one and make the window max observations transparent.
Firstly, create a dual axis chart and synchronize the axis:
From here, you need to make the window max observations transparent and Tableau still maintains the outline of the marks when you hover each different one:
In this way you have now created an easier readable legend for comparison. You can find a packaged workbook that includes the example here: //share.inviso.dk/DualScatter