Hello everyone,
I am trying to show change in a pre-post data set. My visualization is two side-by-side scatter plots with hexagonal binning, one for the pre values and one for the post values. I would like to show change for each entry but keep it within a scatter plot. Is there a better way to show change for each entry? I created the plots with iNZight. Any help is greatly appreciated. Thank you!
If you want to show change, then show just that. Compute the difference or the ratio and plot only this derived variable. Show the magnitude of change and its direction, possibly as if they were two different variables. If the original data must be shown but is not as important you can leave it to an inset or to interaction.
Fernando, thanks so much, your comments really helped me to focus on what I really want to show. John
As Fernando said, creating a ‘change’ variable then plotting it will draw attention to any significant changes that have taken place. There doesn’t seem to be much of a relationship between the EScore and PScore in your data so I’m not sure how useful plotting them against each other is. I’ve gone with a bar chart for each score type showing the value change from pre to post for each ID with a diverging gradient colour scale to highlight the more extreme changes in value. If you’re more interested in % change you could plot that instead of value change.
Attached image:
Paul, many thanks, I really appreciate your idea of displaying the change variable in bar charts. This is a great help. John