Chart Types and Uses

Bar chart

Bar chart best practices

• Quantitative-based
• Proportionate to the data
• Goal: Comparison

Bar chart best practices

• Quantitative-based, easy to read
• Proportionate to the data
• Goal: Comparison
• Avoid: More than 10 comparisons
• Vertical: Easiest to read, good for time intervals
• Stacked: Parts of a whole for each
• Horizontal: Good for long labels, not for time factors

Histogram

Histogram chart best practies

• Bar charts with a twist
• Histogram data is numerical, bar charts are categorical
• With bar charts, each column represents a group defined by a categorical variable; and with histograms, each column represents a group defined by a continuous, quantitative variable.
• Note: No space between bars

Line charts

Line charts best practices

• Quantitative, easy to follow
• Goal: Change over time, particularly financial or historical information
• Avoid: Datasets with little change, more than three categories
• Disadvantage: Only express one trend or idea and usually have to label high(s) and low(s)
• Note: Keep scales consistent

Area charts

Area charts best practice

• Quantitative, easy to follow trend
• Goal: Time-series relationship that includes volume
• Disadvantage: Requires two or more categories with some relationship that can be visualized without conflicting, totals are not as important as relational rise and fall
• Note: Use transparent colors for overlapping data
• Avoid: Too many categories, more than one trend

Scatter plots

Scatter plots best practices

• Ability to visualize clusters, patterns, and relationships in a cloud of data points—especially a very large one
• Advantage: Good for scientific data
• Goal: Correlations (negative or positive especially) and trends
• Disadvantage: Needs larger datasets with relational data
• Note: Can draw lines to emphasize connections

Bubble cloud

Bubble cloud best practices

• Ability to visualize clusters, patterns, and relationships in a cloud of data points—especially a very large one
• Advantage: Good for scientific data
• Goal: Correlations (negative or positive especially) and trends
• Disadvantage: Needs larger datasets with relational data
• Advantage: Size of dots can emphasize data type, scatter plot on steroids

Heat Chart

Heat chart best practices

• Use color to communicate relationships between data values that would be would be much harder to understand if presented numerically in a spreadsheet
• Disadvantage: Not a set standard in representation
• Note: Colors have meaning and should correspond with data representation

Pie chart

Pie chart best practices

• Items must add up to 100% of a whole (be careful of fractions!)
• Percentage-based but actual number totals can provide credibility
• Goal: Breakdown
• Avoid: More than 7 slices, awkward labeling

Donut charts

Donut chart best practices

• Items must add up to 100% of a whole
• Percentage based, but actual numbers give credibility
• Advantage: Can provide series inside of categories
• Goal: Breakdown
• Disadvantage: Can be overwhelming if data isn't relationally clear
• Avoid: More than 5 slices, more than 3 rings, too much info in the middle

Gantt charts

Other data types

• Box plots
• Funnel charts
• Bullet graphs
• Spider charts
• Venn diagrams

Other issues

• Colors for lines/slices/bars
• Font size and color
• AP Style use
• Label choices, lengths and positions

Evaluating chart use

• What do you really need to say or demonstrate?
• What kind of data are you working with?
• What is the most efficient way to do this?
• What can your reader understand?
• What are your resources: time, money, manpower, brainpower, dataset size and quality, tools?