Visual Vocabulary: A Guide to Effective Data Visualization

Financial Times Visual Vocabulary: A Guide to Effective Data Visualization
The Financial Times Visual Vocabulary is a comprehensive resource designed to assist designers and journalists in selecting the optimal symbology for data visualizations. Created by the Financial Times Visual Journalism Team, this guide serves as a cornerstone for newsroom-wide training aimed at improving chart literacy. Inspired by the Graphic Continuum by Jon Schwabish and Severino Ribecca, the Visual Vocabulary is not intended to teach chart creation, but rather to help recognize opportunities for effective chart usage alongside written content.
Deviation Charts: Highlighting Differences
Deviation charts are designed to show how data deviates from a reference point or baseline, making them ideal for highlighting differences, changes, or outliers. These visualizations are particularly useful for understanding variations, financial comparisons, or survey results. By clearly illustrating deviations, these charts help viewers quickly grasp the magnitude and direction of differences from a central point.
Diverging Bar Chart
Bars extend on either side of a central line, showing positive and negative deviations.
Dumbbell Plot
Displays two values for each category, connected by a line to show range or difference.
Lollipop Chart
Combines a bar chart concept with a dot at the end of a line for visual simplicity.
Surplus/Deficit Filled Line
Highlights areas above and below a baseline with different colors for financial data.
Correlation Charts: Visualizing Relationships
Correlation charts are essential tools for visualizing relationships between two or more variables. These charts help identify trends, clusters, or outliers, making them invaluable for exploratory data analysis. By plotting data points based on multiple variables, correlation charts allow viewers to quickly grasp complex relationships within datasets.
Scatter Plot
Plots individual data points based on two variables, showing linear or non-linear relationships.
Connected Scatter Plot
Adds lines connecting sequential data points, indicating flow or order over time.
Bubble Chart
Expands on scatter plots by using bubble size to represent a third variable.
XY Heatmap
Uses color intensity on a grid to represent correlation strength between variables.
Ranking Visualizations: Ordering Data
Ranking visualizations arrange data points in a specific order, either ascending or descending, to facilitate easy comparisons. These charts are ideal for presenting survey results, ratings, or leaderboard-style data. By visually ordering information, ranking charts allow viewers to quickly identify top performers, lowest values, or relative positions within a dataset.
Ordered Bar/Column Chart
Sorts bars or columns based on their values, clearly showing category performance.
Dot Strip Plot
Plots individual points along a line, comparing multiple items in ranked order.
Slope Chart
Shows how rankings change over time by connecting data points with lines.
Bump Chart
Illustrates changes in rank for multiple entities over time, showing performance shifts.
Distribution Charts: Understanding Data Spread
Distribution charts are crucial for showing the spread or variability in a dataset, helping to understand patterns, clusters, and outliers. These visualizations provide insights into the shape of data distributions, revealing important characteristics such as central tendencies, skewness, and data concentration. By illustrating how data is distributed, these charts aid in identifying trends and making informed decisions based on data patterns.
Histogram
Groups data into bins, using bars to show frequency in each bin. Ideal for continuous data.
Dot Plot
Represents individual data points along a scale, useful for smaller datasets.
Boxplot
Displays median, quartiles, and outliers, effective for comparing distributions.
Violin Plot
Combines boxplot with density plot, showing probability distribution across groups.
Change Over Time: Tracking Trends
Visualizations of changes over time are crucial for trend analysis, forecasting, and understanding shifts in data points. These charts help viewers grasp how variables evolve over different time periods, from short-term fluctuations to long-term trends. By effectively illustrating temporal changes, these visualizations enable better decision-making based on historical patterns and future projections.
Line Chart
Displays data points connected by lines over time, ideal for continuous data.
Column/Bar Chart
Uses vertical or horizontal bars to show data changes across time intervals.
Area Chart
Fills the area below the line to emphasize magnitude of changes over time.
Candlestick Chart
Shows open, close, high, and low values, commonly used in finance.
Fan Chart
Illustrates forecasts with increasing uncertainty over time using shaded bands.
Magnitude Charts: Comparing Quantities
Magnitude charts are designed to compare quantities across different categories or time points, showing the relative size or frequency of values. These visualizations are particularly effective when the focus is on understanding the scale of differences between data points. By presenting data in a way that emphasizes size comparisons, magnitude charts allow viewers to quickly grasp which categories or entities are larger, smaller, or similar in scale.
Bar/Column Chart
Uses horizontal or vertical bars to compare quantities among categories, providing a clear visual representation of differences.
Pictogram Chart
Utilizes icons to represent quantities, with each icon equaling a set number of units. Proportional pictograms vary icon sizes to show magnitude differences.
Lollipop Chart
Combines lines and points to indicate the magnitude of each data point, drawing more attention to the data value than standard bar charts.
Part-to-Whole Charts: Visualizing Composition
Part-to-whole visualizations illustrate how individual components contribute to a larger total. These charts are essential for showing the composition of a whole and the relative sizes of its parts. They help viewers understand proportions, percentages, and the overall structure of data. Part-to-whole charts are particularly useful when the relationship between individual elements and their sum is crucial to the story being told with the data.
Stacked Bar/Column Chart
Each segment represents a part of the total, suitable for showing proportions in categories.
Treemap
Uses nested rectangles to represent hierarchical data, with size indicating proportion of the total.
Pie Chart
Shows proportions of a whole, with each slice representing a percentage of the total.
Sunburst Diagram
Similar to a pie chart but with additional layers to show hierarchy within the data.
Spatial Charts: Mapping Geographic Data
Spatial charts add geographic elements to visualizations, showing spatial relationships and patterns. These charts are crucial when precise locations or geographical patterns in data are more important to the reader than other aspects. Spatial visualizations help in understanding regional variations, identifying clusters, and visualizing data in a geographic context.
Choropleth Map
Shades regions based on data values, suitable for showing geographic data distributions.
Proportional Symbol Map
Uses symbols of varying sizes to represent data quantities on a map.
Flow Map
Displays movement or flow between locations on a map.
Contour Map
Shows lines of equal value, often used in topography and weather mapping.
Flow Charts: Visualizing Connections
Flow charts visualize connections or relationships between entities, showing how elements interact or move within a system. These visualizations are particularly useful for illustrating processes, transitions, or the movement of resources. Flow charts help viewers understand complex relationships and sequences, making them invaluable for explaining intricate systems or multi-step processes.
Sankey Diagram
Uses arrows or paths to represent flow magnitudes between categories.
Flow Map
Shows movement between geographic areas on a map.
Chord Diagram
Illustrates relationships between data points in a circular layout.
Network Diagram
Shows connections and interdependencies between entities in a network structure.
Diverging Bar Chart: Highlighting Positive and Negative Values
The diverging bar chart is a powerful tool for visualizing deviations from a central point. It extends bars on either side of a central line, with positive values typically extending to the right or up, and negative values to the left or down. This chart type is particularly effective for displaying sentiment analysis (positive vs. negative opinions) or performance deviations from a target.
Key Features
- Clear central reference point - Easy comparison of positive and negative values - Effective for showing relative magnitudes
Best Uses
- Survey results with agree/disagree scales - Financial performance relative to expectations - Temperature deviations from a norm
Dumbbell Plot: Comparing Two Data Points
The dumbbell plot, also known as a DNA chart, is an excellent choice for displaying the range or difference between two related values for each category. This chart type connects two data points with a line, resembling a dumbbell shape. It's particularly useful for comparing 'before and after' scenarios or showing the change between two time points.
1
Advantages
Dumbbell plots excel at showing the magnitude of change between two values, making it easy to identify significant shifts or stable categories at a glance.
2
Design Considerations
Use contrasting colors for the start and end points to enhance readability. Consider adding data labels for precise values.
3
Application
Ideal for visualizing changes in rankings, comparing survey results over time, or illustrating gaps between actual and target values across multiple categories.
Scatter Plot: Exploring Relationships Between Variables
Scatter plots are fundamental tools in data visualization, used to display the relationship between two continuous variables. Each data point is represented by a dot on the chart, with its position determined by the values of the two variables being compared. This visualization is particularly effective for identifying patterns, correlations, or outliers in datasets.
Key Features
- Reveals patterns and trends - Identifies correlations (positive, negative, or none) - Highlights outliers and clusters
Best Practices
- Use clear axis labels - Consider adding a trend line - Adjust point size or color for additional variables
Applications
- Scientific research - Economic analysis - Performance metrics comparison
Bubble Chart: Adding a Third Dimension
Bubble charts expand on the concept of scatter plots by introducing a third variable, represented by the size of each bubble. This additional dimension allows for more complex data representation, making bubble charts ideal for visualizing datasets with three related quantitative variables. The position of each bubble on the x and y axes represents two variables, while the bubble's size represents the third.
1
Data Preparation
Identify three related variables suitable for x-axis, y-axis, and bubble size.
2
Design Considerations
Ensure bubble sizes are proportional to data values and use a clear scale for interpretation.
3
Interpretation
Guide viewers on how to read the chart, explaining what each axis and bubble size represents.
4
Interactivity
Consider adding hover effects or tooltips for detailed information on each data point.
Ordered Bar Chart: Simplifying Data Comparison
Ordered bar charts are a powerful variation of standard bar charts, where bars are sorted in ascending or descending order based on their values. This simple yet effective organization makes it easy for viewers to quickly identify the highest and lowest values, as well as understand the overall distribution of data across categories. Ordered bar charts are particularly useful when the specific ranking of items is more important than their original categorical order.
Key Benefits
- Instantly reveals highest and lowest values - Simplifies comparison across categories - Highlights distribution patterns
Design Tips
- Use consistent color scheme - Consider adding data labels - Ensure adequate spacing between bars
Common Applications
- Sales rankings - Survey result analysis - Performance metrics across departments
Slope Chart: Visualizing Change Between Two Points
Slope charts are elegant visualizations that excel at showing how rankings or values change between two points, typically over time. Each entity is represented by a line, with the slope indicating the direction and magnitude of change. This chart type is particularly effective for highlighting significant changes, stable entities, and overall trends within a dataset.
1
Structure
Slope charts consist of two vertical axes representing the start and end points, with lines connecting corresponding values for each entity.
2
Interpretation
Steeper slopes indicate more significant changes, while flat or nearly flat lines show stability. Crossing lines represent changes in ranking.
3
Best Practices
Use color coding to highlight specific trends or entities of interest. Consider adding data labels at both ends for clarity.
4
Ideal Uses
Comparing performance metrics before and after an intervention, visualizing changes in market share, or showing shifts in demographic data over time.
Histogram: Understanding Data Distribution
Histograms are fundamental tools for visualizing the distribution of continuous data. They group data into bins (intervals) and use bars to represent the frequency or count of data points within each bin. This chart type is crucial for understanding the shape, central tendency, and spread of a dataset, making it invaluable in statistical analysis and data exploration.
Key Components
- X-axis: Represents data values divided into bins - Y-axis: Shows frequency or count - Bars: Represent data frequency within each bin
Interpretation
- Shape: Reveals distribution type (e.g., normal, skewed) - Central tendency: Indicates typical values - Spread: Shows data variability
Best Practices
- Choose appropriate bin sizes - Ensure consistent bin widths - Consider density plots for smoother representation
Boxplot: Summarizing Data Distributions
Boxplots, also known as box-and-whisker plots, are powerful tools for summarizing and comparing distributions of data. They provide a concise representation of key statistical measures, including the median, quartiles, and potential outliers. Boxplots are particularly useful for comparing distributions across multiple categories or groups, making them invaluable in statistical analysis and data exploration.
1
Components
Box: Represents the interquartile range (IQR), containing the middle 50% of the data.
2
Median Line
Horizontal line within the box, showing the median value.
3
Whiskers
Extend from the box to show the range of the data, typically to 1.5 times the IQR.
4
Outliers
Individual points beyond the whiskers, representing potential outliers in the dataset.
Line Chart: Tracking Changes Over Time
Line charts are one of the most common and versatile tools for visualizing changes over time. They connect individual data points with lines, making it easy to see trends, patterns, and fluctuations in continuous data. Line charts are particularly effective for showing how one or more variables change over a continuous interval, such as time series data.
Key Features
- X-axis typically represents time - Y-axis shows the measured variable - Lines connect data points to show trends - Multiple lines can compare different series
Best Practices
- Use clear, contrasting colors for multiple lines - Consider adding data points for precise values - Use appropriate scale to highlight important changes - Add a legend for multiple series
Applications
- Stock price movements - Temperature changes over time - Sales trends across months or years - Website traffic analysis
Area Chart: Emphasizing Volume Over Time
Area charts are a variation of line charts where the area between the line and the x-axis is filled with color or shading. This visualization is particularly effective for showing cumulative totals over time or emphasizing the magnitude of changes. Area charts can be used to display a single series or multiple series stacked on top of each other, making them versatile for various data storytelling needs.
1
Single Area Chart
Ideal for showing the change in a single variable over time, emphasizing its magnitude.
2
Stacked Area Chart
Shows how multiple variables contribute to a total over time, with each variable stacked on top of the others.
3
100% Stacked Area Chart
Displays the relative percentage each variable contributes to the total at each point in time.
4
Design Considerations
Use transparent colors to prevent lower layers from being obscured in stacked versions. Consider using gradients for aesthetic appeal.
Pie Chart: Visualizing Part-to-Whole Relationships
Pie charts are circular graphs that display data in slices, where each slice represents a category's proportion of the whole. While often debated in data visualization circles, pie charts can be effective for showing simple part-to-whole relationships, especially when there are few categories and the differences between them are significant. They are particularly useful when the focus is on comparing parts to the whole rather than parts to each other.
Strengths
- Intuitive for showing percentages - Effective for few categories (ideally 3-5) - Good for highlighting dominant category
Limitations
- Difficult to compare slice sizes accurately - Less effective for many categories - Can be misleading if not carefully designed
Best Practices
- Order slices from largest to smallest - Use clear, contrasting colors - Add percentage labels for clarity - Consider a donut chart for a modern look
Treemap: Hierarchical Part-to-Whole Visualization
Treemaps are powerful visualizations for displaying hierarchical data using nested rectangles. Each rectangle's size is proportional to the quantity it represents, making treemaps excellent for showing both the hierarchy and the relative size of categories simultaneously. This chart type is particularly useful when dealing with large datasets that have multiple levels of subcategories.
1
Structure
Rectangles are nested within larger rectangles, representing hierarchical relationships.
2
Size
The area of each rectangle is proportional to the quantity it represents.
3
Color
Can be used to represent an additional variable or to distinguish between categories.
4
Interactivity
Often includes zoom or drill-down features for exploring subcategories.
Choropleth Map: Visualizing Geographic Data
Choropleth maps use color-coding to display statistical variables across predefined geographic areas. These maps are excellent for showing how a variable changes across regions, such as countries, states, or counties. By using different shades or intensities of color, choropleth maps can effectively communicate variations in data values across a geographic space, making them invaluable for spatial analysis and regional comparisons.
Key Features
- Color-coded regions - Legend explaining color scale - Clear geographic boundaries - Data values typically represented as rates or ratios
Best Practices
- Use a sequential or diverging color scheme - Ensure color scale is intuitive and colorblind-friendly - Normalize data by population when appropriate - Include a clear legend and title
Applications
- Population density - Election results - Economic indicators by region - Health statistics across areas
Flow Map: Visualizing Movement and Connections
Flow maps are specialized geographic visualizations that show the movement of objects, people, or information between different locations. These maps use lines or arrows of varying thickness to represent the volume or intensity of flow between points. Flow maps are particularly effective for illustrating migration patterns, trade routes, or any data that involves movement between distinct geographic locations.
1
Key Components
Lines or arrows connecting locations, with thickness representing volume of flow.
2
Color Coding
Can be used to distinguish different types of flow or to indicate directionality.
3
Base Map
Typically a simplified geographic map showing relevant locations or regions.
4
Data Representation
Flow volume can be shown through line thickness, color intensity, or both.
Sankey Diagram: Visualizing Flow and Transfers
Sankey diagrams are flow diagrams where the width of the arrows or streams is proportional to the flow quantity. These diagrams are excellent for visualizing the transfer of energy, materials, or costs between processes. They help in understanding complex systems by showing how resources or quantities are distributed across various stages or categories, making them invaluable for process analysis and resource allocation studies.
Structure
- Nodes represent stages or categories - Streams show flow between nodes - Width of streams indicates quantity
Color Coding
- Can represent different types of flow - Helps in tracing specific streams through the system
Best Practices
- Minimize stream crossings for clarity - Use consistent color scheme - Include clear labels and a legend
Network Diagram: Visualizing Complex Relationships
Network diagrams, also known as node-link diagrams, are powerful tools for visualizing relationships and connections between entities. These diagrams represent entities as nodes (points or circles) and relationships as edges (lines connecting the nodes). Network diagrams are particularly useful for showing social networks, organizational structures, or any system where the connections between elements are as important as the elements themselves.
1
Nodes
Represent entities or data points. Size or color can indicate importance or category.
2
Edges
Show connections between nodes. Can be directed (with arrows) or undirected.
3
Layout
Arrangement of nodes can reveal clusters, central figures, or isolated elements.
4
Interactivity
Often includes zoom, pan, and click features for exploring complex networks.