# An appropiate data visualization?

I want to include a plot in my thesis to highlight regional differences in bicycle collision outcomes. Now there are several plot visualizations I could use, and I am not sure which one could be the most appropriate. Having a "scale free" y axis better highlights the differences between each region, but it may also confuse the reader.

I am aware that there might not be an optimal solution to this, but I would appreciate to hear your thoughts.(If you have a completely other suggestion for an appropriate plot than the ones below, I would also be happy to hear those).

These are the plots I have made:

Plot 1 Plot 2 Plot 3 Plot 4

Looking into so many plots may miss the bigger picture.
We should keep the initial plot in "One Consolidated plot" e.g.

Can add few other consolidated views - Pie on aggregated data on region and Vehicle

• Whenever you give percentages, you need a very clear statement about the denominator. I can deduce it here, but in general one could give the percentage of bike only accidents of all accidents in the medium sized cities or the percentage of bike only accidents happening in medium sized cities of all bike-only accidents. Or...

• The kind of percentage (or absolute numbers) needs to fit your analysis purpose - we cannot tell this without knowing more details about your analysis.
Without details, I'd report the absolute numbers (or accidents per distance biked or per time on bike).

• I'd go for the same y axis for all facets.

• I'd try out a scatter plot instead of the bar plot. Here, that would be parallel coordinate plot. A polar variant of parallel coordinate plots are star plots.

• Regardless of the type of plot (parallel coordinates, star, bar) you have two grouping variables (region type and accident type). Try out both ways: region as x/angle axis and accident as color/facet and accident type as x/angle with region type color/facet. I'd typically choose one diagram for the main text and put the other into the supplementary material or appendix.

• If you choose to put a diagram of percentages into your text, consider putting also the diagram for the absolute numbers into the supplementary material or appendix.

In my opinion, given the complexity of your information, the plot that best achieves your goal without being misleading is plot 4.

Even though your Y-axis changes, all the info is very clear to the reader.

I think the important choice here is if you want to show them the clearest way or if you you want to lead them to other conclusions, using other plots.