So I sent this in broad strokes to someone at work and thought it’d be good to share more generally.
So if you’re keen on my thoughts on the data vis, space check it out.
Here are a few general principles I like to follow for information visualisation.
3–30–300 Rule
The 3–30–300 rule for information visualisation suggests that people can comprehend information most effectively when presented in three seconds, thirty seconds, or three minutes intervals.
Specifically, it states that:
- In the first 3 seconds, an information visualisation should capture the viewer’s attention and convey a clear message or purpose.
- In the next 30 seconds, the viewer should be able to grasp the main points or message of the visualisation.
- In the following 3 minutes, the viewer should be able to understand the details and nuances of the information presented.
The story chooses the graph — sketch and reflect
There is always an optimal chart for the story and dataset you’re working with.
Sometimes it’s great to just sketch something out to give your team and clients a decent idea of the way you’re thinking about the problem.
A picture tells 1000 words. Besides, sketching is fun. The creativity wears you out at times but it’s worth the effort to get feedback.
For example, here are some good rules of thumb for the 10 most common data visualisation formats
- Bar chart: Use horizontal bars for long category labels or when comparing a large number of categories, and vertical bars for data that are time-based or involve numeric values.
- Line chart: Use solid lines for continuous data and dashed lines for predicted or estimated data, and label axes clearly.
- Scatter plot: Use a regression line or trend line to help identify patterns, and consider adding reference lines to show important thresholds or benchmarks.
- Pie chart: Avoid using more than 5–6 categories, and consider using a bar chart or stacked bar chart instead if there are more categories.
- Heatmap: Use a diverging color scale to highlight extremes in the data, and consider using a dendrogram to show hierarchical clustering.
- Treemap: Use a color scheme that highlights important categories or subcategories, and use labels or annotations to help clarify the meaning of the different areas.
- Network diagram: Use a layout that emphasizes the most important nodes or connections, and consider using different node shapes or colors to represent different types of nodes or connections.
- Geographic map: Use a projection that minimizes distortion and accurately represents the areas and shapes of different regions, and use different colors or symbols to represent different values or categories.
- Bubble chart: Use a color scheme that helps distinguish between different groups or categories, and use labels or tooltips to provide additional information.
- Choropleth map: Use a color scheme that accurately represents the data and avoids misleading patterns, and consider using a log scale or other transformations to help show differences across different regions.
Resources
Here are some resources that might be helpful for inspiration. Especially when sketching:
- Data Visualization Society: This is a global community of data visualization professionals that offers resources, discussions, events, and networking opportunities.
- FlowingData: This website provides tutorials, data visualization tools, and news related to information visualization.
- Visualising Data: This site offers tutorials, resources, and examples of information visualization projects.
- Information is Beautiful: This is a blog and data visualization agency that showcases creative and innovative data visualizations.
- Storytelling with Data: This website provides tips, tutorials, and examples of how to effectively use data visualization to tell a story.
- The Data Visualisation Catalogue: This is an online library of different types of data visualization with explanations of their strengths and weaknesses.
- Visual Capitalist: This website specializes in creating data visualizations that tell stories about business and economics.
And some links for these:
- Data Visualization Society: https://www.datavisualizationsociety.com/
- FlowingData: https://flowingdata.com/
- Visualising Data: https://www.visualisingdata.com/
- Information is Beautiful: https://informationisbeautiful.net/
- Storytelling with Data: https://www.storytellingwithdata.com/
- The Data Visualisation Catalogue: https://datavizcatalogue.com/
- Visual Capitalist: https://www.visualcapitalist.com/
