In data visualization, where information speaks through visual cues, the choice and application of colors play a vital role in conveying insights effectively. And consistent color coding is a critical element in ensuring clarity. It helps quickly grasp patterns and identify categories. By doing this inconsistently, we might suggest the connections between unrelated elements.
An example of such a case is the WindEurope chart included in the Reuters article. The graph presents the installed wind capacity in Europe at the end of 2022. It consists of the pie chart showing the share of the total wind capacity by country and the stacked bar chart showing the distribution of the location type within each country. The layout selection blurs the key message and requires a back-and-forth jumping between two charts. Additionally, the color coding is misleadingly links’ Germany’ with ‘onshore’, and ‘Spain’ with ‘offshore’. We could switch colors to avoid duplicates, but the better alternative is to rethink the layout.
Elements that work in this chart
✔ Different perspectives
Although using the pie chart may be unfortunate, it brings valuable information that enriches the analysis. While the bottom chart allows us to gain insights into the actual production levels of wind energy across various European nations, the pie chart depicts the proportionate contribution of each country’s wind capacity in relation to the total European wind capacity.
✔ Grouping the countries
Simplifying the layout by grouping countries with smaller wind capacity contributions enables us to maintain the full context while directing our attention toward the most impactful countries.
Elements that don’t work in this chart
✗ Misleading color usage
Encoding different data with the same colors creates a misleading connection between unrelated elements. In the pie chart, dark blue means’ Germany’ while on the column chart, it means ‘onshore’. Similarly, light blue means either ‘Spain’ or ‘offshore’. This practice not only creates a potential for confusion but also requires unnecessary cognitive effort.
✗ Bars orientation
When employing a column chart with lengthy category names, it becomes necessary to reduce the font size and, in some cases, even split the names into two lines. This makes the categories harder to read and the text weirdly aligned.
Step-by-step improvements
✎ Flip the chart
We should start by transitioning from a column chart to a bar chart. By doing so, we gain additional room for the labels, allowing us to increase the font size and accommodate them within a single line. This improves readability by aligning with natural reading patterns.
✎ Remove the misleading color coding
Since both charts display the same level of detail, resolving the color-coding issue can be easily achieved by switching the chart type and integrating the data into the existing chart structure. This approach simplifies the layout and eliminates any inconsistency in grouping countries under the ‘other’ category (in the pie chart, there are eight more countries compared to the column chart).
✎ Adjust chart type
By opting for a panel bar chart instead of a stacked bar chart, we can introduce diverse viewpoints and enhance the depth of analysis. In this format, the first column represents the overall value expressed as a percentage, while the two subsequent columns display the actual capacities in gigawatts (GW) split by the category (onshore and offshore). This approach ensures a consistent baseline across all bars while preserving the total capacity information.
✎ Work on the formatting
We can improve readability by applying alternating colors to every second row. This simple technique ensures the immediate visibility of the connection between country names and their corresponding bars. Lastly, optimizing the formatting of headers and eliminating unnecessary elements further improves the chart’s overall clarity.
