Skip to Main Content
It looks like you're using Internet Explorer 11 or older. This website works best with modern browsers such as the latest versions of Chrome, Firefox, Safari, and Edge. If you continue with this browser, you may see unexpected results.

Data Visualization

Bar chart

: bar chart showing female and male occupations from the 1881 census of England and Wales.

Figure 4: Example of a bar chart.
Smith, V. (1996). Male and Female Occupations From the 1881 Census of England and Wales.
CC BY-SA 4.0

Best used for:

  • Reading numbers accurately from the chart
  • Comparing individual bars to each other
  • Exploring overall trends across categories
  • Note, see histograms for continuous data.;

Data

  • One (or more) text (categorical) variable
  • One numerical variable

Strength

  • Familiar to most audiences
  • High perceptual accuracy because they use alignment and length

Best practices

  • Use horizontal bars to enhance label readability
  • Consider sorting bars by length when alphabetical order is not important
  • Axis must start at zero

How do I create a bar chart?

Histogram chart

Histogram showing the frequency of specific heights of cherry trees in feet.

Figure 5: Example of a histogram.
Unknown. (2008). Heights of Black Cherry Trees
CC BY-SA 3.0

Best used for

  • To show the density and distribution of continuous numerical data.  Note, that bar charts plots categorical data

Data

  • One continuous numerical variable

Strength

  • High perceptual accuracy because they use alignment and length

Best practices

  • Divide the entire range of values into a series of intervals (bins) and then count how many values fall into each interval. 
  • The bins are usually specified as consecutive, non-overlapping intervals of a variable.
  • The bins (intervals) must be adjacent, and are often (but are not required to be) of equal size 

How do I create a bar chart?

Pie chart

Pie chart showing 4 unlabelled categories.
Figure 6: Example of a pie chart.
Sylvanmoon. (2016). Pie-Chart.
CC0

Best used for

  • When separate categories add up to a meaningful whole

Data

  • One text (categorical) variable
  • One numerical variable

Strength

  • Familiar to most audiences

Best practices

Pie charts are very difficult for humans to interpret accurately due to angle and area.  To improve perception:

  • Categories MUST add up to meaningful whole
  • Only use when significant differences exist in category size
  • Sort by size
  • Ideally, limit your categories to two: the data point you wish to highlight and everything else.
  • Avoid 3D versions as they cause distortion.

How do I create a pie chart?

Line chart

 Line chart showing interest rates by compounding frequency

Figure 7: Example of a line chart
Kazov. (2015). Compound Interest (English).
CC BY-SA 4.0

Best used for

  • Visualizing continuous and temporal data

Data

  • One date (categorical) variable on the x-axis (horizontal)
  • One or more numerical variable on the y-axis (vertical)

Strength

  • Familiar to most audiences
  • Clearer choice over grouped and stacked bar charts when multiple categories are needed

Best practices

  • Use colour to differentiate categories
  • Too many lines creates difficulty in readability
  • Place labels to improve readability
  • Add small dots to clarify measurement locations
  • Axis must start at zero

How do I create a line chart?

Scatter plot

Scatter plot showing the correlation between oppose and support votes

Figure 8: Caption: Example of a scatter plot.
Opabinia regalis. (2015). Scatter Plot of FfA votes on English Wikipedia.​
CC BY-SA 4.0

Best used for

  • Showing how one variable affects the other, or for showing precise, data dense visualizations, correlations, and clusters between two variables.

Data

  • Two numerical variables  
  • When a third numerical variable can be added to create a bubble chart.  The third variable changes the size of the dots. 

Strengths

  • High perceptual accuracy
  • Visual representation of relationship between variables including:
  • Show pos / neg correlation, clusters, outliers, 

Best practices

  • Scatter plots are more commonly used in the social and hard sciences.  It may be necessary to provide training for your audience for accurate interpretation.
  • Use transparency or aggregation for stacked dots
  • Show variables requiring most precision on axis, 3rd variable to encode size

How do I create a scatter plot?

Heat map

Heat map showing high and low percentages in individual cells.

Figure 9: Example of a heat map.
LuckyDawg10. (2017). Mike Zunino Swing% vs All Pitchers.
CC BY-SA 4.0

Best used for

  • Showing general trends

Data

  • Data grid of two numerical or categorical variables
  • Third variable is (often the number of data points associated with the particular row and column) is encoded as the colour of the cell.

Strength

  • Useful for representing large amounts of data
  • Space efficient

Best practices

  • Heat maps are not commonly used making it difficult for most people to read.  Provide training for your audience for accurate interpretation.
  • Do not use if numerical value needs to be read precisely
  • Use colour encoding to give a sense of high and low values
  • Heat maps are highly dependent on the order of the rows and columns to reveal similarities between different categories within the same variable

How do I create a heat map?

Maps - choropleth

Choropleth map of the United States showing 6 categories of vehicles per 1000 people.

Figure 10: Example of a choropleth map.
Bramo126. (2017). Map of each state's vehicles per 1000 people in 2015
CC BY-SA 4.0

Best used for

  • When spatial patterns are important

Data

  • Place variable (x,y; longitude and latitude; address; place name)
  • One numeric variable or one categorical variable

Strength

  • Easy to read and create
  • Data points do not overlap
  • Recognizable shape of geographic space does not require labels

Best practices

  • Be aware of how region size may confuse message (ie: elections - vast northern regions vs condensed cities)
  • Assess if spatial data is better displayed via graphs

How do I create a choropleth map?

Maps - symbol

Map of Africa showing proportional dots for fertility and colour coded countries for GDP in US dollars.

Figure 11: Example of a symbol map.
GraceN.Cartography. (2016). Adolescent Fertility and Poverty in Africa.jpg
CC BY-SA 4.0

Best used for

  • When spatial patterns are important

Data

  • Place variable (x,y; longitude and latitude; address; place name)
  • One or two numeric variable or categorical variable

Strength

  • Easy to read and create
  • Can encode two separate variables in addition to location
  • Best used for cities

Best practices

  • Use transparency or aggregation for overlapping circles
  • If precision is important, use bar charts instead

How do I create a symbol map?

Suggest an edit to this guide

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.