Recognize the Importance of Color, Icons, and Shapes
Learning Objectives
After completing this unit, you’ll be able to:
- Show how to use colors, shapes, and icons with awareness and intention.
- Explain the impacts and implications of colors, icons, and shapes in your data visualizations.
Use Color with Sensitivity and Awareness
Just as you must carefully curate the language you’re using, you must also make informed choices about the use of color for your data visualizations. Good color palettes for data visualization should, at minimum, meet basic accessibility guidelines and offer sufficient contrast between colors for readers with vision difficulties.
Going beyond accessibility, color choices should also avoid reinforcing gender or racial stereotypes, such as baby pink and baby blue to represent women and men, or colors associated with skin tones or racial stereotypes (for example, black to represent Black people, yellow to represent Asian people).
The legend below shows a problematic color scheme applied to data on race and ethnicity. The red aligned with minority groups such as Black or Hispanic can have negative connotations in Western culture—often associated with danger or aggression.
Color can also imply a hierarchy, and perpetuate harmful stereotypes. For example, using a series of red shades for groups of color alongside a distinct blue color for white groups creates a visual divide, or even competition, between the two. Also, a graduated color palette should not be used for categorical data, as it shows greater or higher values in darker colors and smaller or lower values in lighter colors.
There are also emotional connotations associated with certain hues. In Western cultures, colors such as red can be perceived as threatening or aggressive and therefore can paint the population you’re visualizing in a negative light. Avoid situations where participants who are depicted in the data visualization may feel they are presented as a problem. While it may be difficult to use separate colors for each category, there are many free color tools available for designers and developers.
When used appropriately, color is a powerful aid in data visualization. Lacking awareness, you risk creating misperceptions and offending your audiences. Again, it’s important to ask yourself: If I were one of the data points on this visualization, would I feel offended?
Understand the Power of Iconography
Data visualizations rely on the use of pictures, icons, and shapes. And by their very nature, icons are intended to convey broad meanings. This is both incredibly useful and potentially quite harmful.
Always be mindful about how you depict groups of people. You want to reflect society as a whole by using a mix of genders, races, ethnicities, ages, and other characteristics. Consider these factors by asking a few questions: Who is going to see your results? How might they perceive the icons? How can you avoid perpetuating harmful and offensive stereotypes? Not all icons will correspond to the content you’re presenting.
Mis- or underrepresentation of certain groups in the imagery and iconography used in your data visualizations can be viewed as a failure to have an empathetic approach to racial and gender equity. For example, in job search results, men are often overrepresented in online imagery compared with women. How you employ images represents an opportunity to transform this pattern and push back on forces of oppression.
Similarly, the choice of icons or imagery to represent racial or ethnic groups can play into negative stereotypes. This includes depictions involving poverty, culturally misapplied tropes, or reinforcing traditional power hierarchies (such as, a white male supervisor with a person of color as a subordinate). Instead, images should always represent people as active and empowered and reinforce their dignity, agency, and humanity.
Soliciting feedback can help you make sure your icons are not offensive, and identify any problematic cultural associations that were missed. Also, using shapes instead of icons (such as a series of circles or squares) can work well if the goal is to show individual units of data.
Another factor to consider when using icons or shapes that resemble people and communities is when they might imply something untrue about the data. An image can have multiple meanings, so when using icons, familiarize yourself with the different connotations and take them into account. This could mean stating your exact intentions so that the meaning is clear to the viewer.
Resources
- Video: Do No Harm Guide
- Video Transcript: Do No Harm Guide
- Website: Tableau Do No Harm
- PDF: Diversity, Equity, And Inclusion In Data Visualization: General Recommendations
- PDF: Racial Equity in Data Visualization Checklist
- Book: Algorithms of Oppression: How Search Engines Reinforce Racism, by Safiya Noble