Data Revisualizations

Personal edits of published charts.

Data journalism adheres to the constraints of a brand aesthetic as well as the constraints of its data.
With all data re-visualizations, I attempt to adhere to the aesthetic of the original publisher.
When a fanciful aesthetic comes to mind, I create additional revisions using a more ~artistic~ design.

I would gladly discuss any of these edits, my thought processes, and possible improvements. I am currently creating, and will be adding more revisualizations in the near future. Enjoy!

Edits:
- Removed truncated axis

- Replaced tick marks with opaque & translucent bars to intuitively indicate that the numbers are %s of a whole (as opposed to integers with a unit)

- Included fewer countries/groups of countries
   + Revision 1
        All countries in the EU28 are removed, the EU28 remains visible for comparison. This requires removing Portugal the leading outlier. Because Brazil and Portugal have similar %'s I felt comfortable with this less traditional edit, knowing that the full difference between the "best" and "worst" countries for gender equality in research publishing was represented.
     + Revision 2
       All countries except for the two "best", one "worst", and the data point I assumed was most relevant to the Economist's audience (EU28) are shown. This creates a cleaner graph that is less likely to cause a reader's attention to slip before they can process the general snapshot of the data being shown.
Edits:
- Changed colors of lines to match colors associated with the social media platforms with which they correlate.

- Un-truncated the y-axis. Added redundant '%' signs to each tick label.
     + I have a bias towards including in full any axis that depicts %'s, as can be seen in the revisualization above as well. This is because, after working as a data designer at xtown.la and user testing my own visualization, I believe that a significant portion of data journalism consumers tend to skim graphs and misread axes. %'s represent data points as a part of whole, which I like to remind me viewers by including redundant '%' symbols and always including the full axis. I never want my audience to view a graph and assume that numbers that correspond to '80%' on the y-axis are actually '100%', as the original graph allows.

- Adding line-type redundancy to forecasted data.
   + Depicting collected data and predicted data using the same line can be dangerous. Again, I believe that a fair portion of data vis consumers skim graphs, looking at general shapes and colors before (or worse: without) bothering to study the axes of a graph. Double (or triple) encoding important details is a good way for a data journalist to avoid misleading their audience.

- Adding intuitive color to each graph's title.
   + Adding a bright-to-dark green color scheme to the labels of eachs graph is a visual way to remind the audience that these graphs are separated into a progression of various ages. This double encoding not only solidifies the visual grid of the graphic (shout-out to design principles!) but also draws attention to the graphs' labeling, which is lost under the graphic's title in the original version.

Back to Projects