Quantitative data are useless without interpretation. Data visualizations synthesize the meaning of raw data into coherent insights. When designers prioritize persuasive imagery over accuracy, visualizations are deceiving. To communicate data with integrity, designers must avoid common data visualization mistakes.
If you torture the data long enough, they will tell you everything.John W. Tukey
John Wilder Tukeywas a man dedicated to data. A founding member of Princeton's statistics department and inventor of the termSoftware, Tukey's favorite aspect of analytics was "taking boring, flat data and bringing it to life through visualization". But for all his numerical zeal, Tukey was aware of the ways in which data can be misinterpreted, even warning: "Visualization is often used for evil."
The dual potential for good and evil isn't just limited to data visualization, but given the paradox of today, it's a pressing design consideration. Information is richer and more accessible than ever before, yet governments, media and corporations are pervasivedistrusted.When organizations (intentionally or not) publish misleading visualizations, the trust gap widens.
Which design factors make visualizations deceptive and howDesignerconvey the meaning of data with the utmost clarity?
Blind spots in data visualization
"Graphic excellence is that which conveys the most ideas to the viewer in the shortest amount of time with the least amount of ink in the smallest amount of space."—Edward R. Tufte,The visual representation of quantitative information
Human vision and recognition are among the most incredible phenomena in nature:
- Light catches the eye.
- The lens sends information from the light to the retina.
- The retina translates the information and fires signals through the optic nerve.
- The optic nerve transmits 20 megabits per second to the brain.
The leap from seeing to thinking is instantaneous, and the brain, saturated with physical demands and external stimuli, must conserve energy by prioritizing what to decipher and what to ignore.
In this rapid transition of seeing and understanding,data visualizationsprove their worth.Here, many visualizations tell viewers what they “should” see in the data, and the overworked brain nods in agreement. Confirmation bias prevails. Objectivity is lost.
To be fair, misleading visuals aren't always the by-product of bad intentions, but even honest mistakes misinform viewers. Eyes are impressionable, and people tend to gloss over information in search of quick takeaways.Seeing and recognizing must play a central role in the design of everythingdata visualizations.
10 data visualization mistakes to avoid
1. Misleading color contrast
Color is one of the most compelling design elements. Even subtle color variations trigger strong emotional responses.In data visualization, high color contrasts can lead viewers to believe that the value differences are larger than they really are.
For example, heat maps represent the magnitude of values with color. High values appear orange and red, while lower values appear blue and green. The difference between the values may be minimal, but the color contrast creates the impression of heat and increased activity.
Designer Takeaways:
- Color is more than just a way to differentiate between data series.
- High-contrast color pairings cause viewers to perceive greater data differences.
2. Improper use of 3D graphics
Two-dimensional representations of three-dimensional space have fascinated viewers for centuries, but 3D graphics pose two serious problems for data visualization.
Occlusion occurs when one 3D graphic partially blocks another.It is the result of mimicking the space in the natural world - where objects have different X, Y and Z coordinates. In data visualization, occlusion obscures important data and creates false hierarchies where unobstructed graphics appear most important.
Distortion occurs when 3D graphics recede into or protrude from the image plane through foreshortening.while drawing,shorteningmakes objects appear as if they inhabit a three-dimensional space, but in data visualization it creates even more false hierarchies. Foreground graphics appear larger, background graphics appear smaller, and the relationship between data series is unnecessarily distorted.
Designer Takeaways:
- 3D graphs are appealing, but they have the potential to obscure important information and confuse scale relationships between data series.
- Unless 3D graphics are essential, visualize data in 2D.
3. Too much data
It's a timeless design problem – what to include and what to cut back to communicate clearly. Data visualization is no exception, especially when data is plentiful and thought-provoking.
The temptation? Make a bold impression with a single visualization.
The problem? Humans are not well equipped to calculate the meaning of multiple values abstracted in visual form.
When visualizations contain too much data, information becomes overwhelming and data blends into a graphical soup that most viewers can't stomach.
Designer Takeaways:
- Information overload applies to data visualization. If too much is presented at once, viewers will avoid it.
- It can be more effective to communicate data with multiple visualizations.
4. Omission of baselines and clipping of scale
The data varies, sometimes greatly, as in measuring income levels or voting patterns by geographic region. Designers can manipulate scale values in charts to make visualizations more dramatic or aesthetically pleasing.
A common example isomitting the baselineor start the y-axis somewhere above zero to make the data differences more prominent.
Another example isTruncation of the X valuea data series to make it appear comparable to inferior series.
Designer Takeaways:
- Aesthetic appeal is secondary to accurate data presentation.
- Omitting baselines and truncating scales to intentionally exaggerate or minimize data differences is unethical.
5. Biased Textual Descriptions
The act of suggestion is the art of persuasion.Tell someone what you want them to see in a picture and they probably will.The text that accompanies the visualizations (supporting texts, titles, captions, captions) is intended to provide viewers with objective context and not to manipulate their perception of the data.
Designer Takeaways:
- Biased text often appears when plotting correlations between datasets (and impliescausation).
- Oftentimes, biased texts come from customers, and it's up to the designers to report the issue.
6. Choosing the wrong visualization method
Each data visualization method has its own use cases.For example, pie charts are used to compare different parts of a whole. They are good for budget breakdowns and survey results (same pie) but are not intended for making comparisons between separate datasets (different pies).
A pie chart could be used to visualize the earnings of three competing companies, but a bar chart would make differences (or similarities) between the companies clearer. If the visualization should show sales over time, a line chart would be a better option than a bar chart.
Designer Takeaways:
- Data visualization methods are not one size fits all.
- Knowvariablesthat visualizations must communicate.
7. Confusing correlations
Visualizing correlations between datasets is a helpful way to give viewers a broader understanding of a topic. One way to show correlations is to overlay datasets on the same graph. When correlations are carefully considered, overlays lead to aha moments. If the number of overlays is too large, it will be difficult for viewers to make connections.
It is also possible to visualize correlations in a way that falsely implies causality.A famous example is the association between increased ice cream sales and an increase in violent crime when both are due to warm weather.
Designer Takeaways:
- It can be helpful to highlight correlations with multiple visualizations that are in close proximity. This allows viewers to assess the data and still create connecting links.
- It is worth repeating this. Correlation is not the same as causation.
8. Zooming in on favorable dates
Data and time are inseparable.It is possible to zoom in on timeframes and view dates that positively impact broader narratives.Visualizing financial performance is a common culprit. Imagine a chart showing strong numbers over a short period of time, making it appear as if a company is thriving. Unfortunately, zooming out reveals that the company only enjoyed a small bounce amid a sharp and prolonged decline.
Designer Takeaways:
- If magnified visualizations don't match what the data says as a whole, let viewers know.
9. Avoid common visual associations
Visual design elementsaffect human psychology.Icons, color schemes, and fonts all carry connotations that affect the viewer's perception.When designers ignore these associations or eschew them in favor of creative expression, things seldom fare well.
Analyzing data visualizations is mentally taxing. At the critical moment of perception, the brain may not need the time to decode the reinterpreted meaning of familiar design elements.
Designer Takeaways:
- There are countless opportunities for creative experimentationdata visualization. Don't distract viewers from the data by forcing them to reinterpret common visual associations.
10. Use data visualizations in the first place
data visualizationsgive shape to numbers that are difficult to contextualize. They unmask meaning when data is complex and multiple variables are involved. But visualization is not always necessary.
If data can be communicated clearly and concisely with a statistic, then it should be.If a textual description proves insightful and the representation of the dataform has little impact, a visualization is not necessary.
Designer Takeaways:
- Data visualization is a communication tool. As with all tools, there are times when it is appropriate and times when another tool is more appropriate.
Visualize data with objectivity
There is a tendency to use data visualizations as irrefutable evidence."We have the data. That's what it means. End of the story." But the great scientific minds of the 20th century loved uncertaintyand embraced the fact that even the most compelling data is prone to error.
data visualizationsare not truth claims.They are analytical snapshots—numerical realities fashioned into forms that the human eye can comprehend. WhenDesignerStripping away from embellishments, visualizations cast data in the warm glow of objectivity, dispelling fears of bias and deception.
Let us know what you think! Please leave your thoughts, comments and feedback below.
• • •
Further reading in the Toptal Design Blog:
- The best examples of data visualization and dashboard designs
- The Mind's Eye - A look into the psychology of data visualization
- Data Visualization - Best Practices and Basics
- A complete rundown of the best data visualization tools
- The Complete Guide to UX Research Methods