Blog Post 9

Link to Flourish Diagram:


For my research question, I wanted to study the relationship between free and enslaved, female and male, 14 to 25 years of age and from the 1820s dataset.  I thought if I put any more values/labels, it would be information overload.  If I wanted to study another aspect of this dataset I would need more graphs.  

I had quite a bit of trouble deciding what type of diagram I wanted to use.  On the x-axis I have the states and the y-axis has the population.  At first, I chose a column chart (grouped), but the bars were too small.  There were states I did not want to include because you couldn’t see low population values.  So to fix this issue, I clicked through the various chart types, and I liked column chart (stacked %).  What I like about this diagram is how it shows the percentage.  In addition, it emphasizes the contrast between states with high percentages of free and enslaved.  I did some other details to make it more pleasant to look at like spacing the bars.  Also for appearance, I picked the color palette ‘paired’, making ‘free’ and ‘enslaved’ better to visualize.   

Although it was the best choice for my question, I am still not completely happy with the result.  It does this weird overlap and when I run my mouse over the bars and the values change.  I actually thought the plot was wrong.  For example, when you put your mouse over New York, it shows the dark green as 11, but the light green bar says 210.  So I thought that doesn’t make sense if dark green bar has more percentage.  The bars may be misleading too because the low population in states like Ohio and Vermont.  Hopefully, the audience will understand it is percentage and look at the values, this could be a case where I would need more than one diagram.  

Couple other things I noticed, the audience might have to do math in their head to figure out the percentages.  Especially if the 4 labels are stacked, it is not as clear compared to other visualization plots.  I have never done percentage plots before and from this experience, percentage column charts are more for visualization than getting accurate numbers. 

One thought on “Blog Post 9

  1. This is such a cool visualization! I also think your decisions behind choosing a few pieces of data (rather than overload it) makes this graph very effective to convey this information. Knowing your challenges with the column graph, I think this stacked chart was the right decision (the different colors create a nice distinction too). I understand your concerns, however, with the cursor-hovering issue and having different numbers appear. I don’t know I don’t know if this was an intentional feature or a glitch in Flourish, but I don’t think it takes away from the essence of your chart. From a percentage standpoint, I like how the totality of a column is broken into different colored portions that show different groups of data as part of the whole. Excellent work and observations on this assignment!

Leave a Reply