Data Analytics: Foundations to Practice · Data Visualization Principles
Avoiding Visual Distortion
A chart can display technically accurate numbers while still creating a visually false impression, and this happens through a small, recurring set of design choices worth knowing by name so they can be spotted and avoided.
A bar chart with a y-axis that doesn't start at zero can make a genuinely small difference between two bars look dramatically larger than it actually is, since bar length is the visual signal a viewer instinctively relies on, and a truncated axis breaks the direct, proportional relationship between that visual length and the actual underlying value it's supposed to represent; this specific distortion occurs more often through a careless default software setting than through any deliberate intent to mislead, but it still meaningfully misleads regardless of intent.
- A truncated (non-zero) y-axis breaks the proportional relationship between bar length and actual value, making small differences look dramatically larger.
- Inconsistent axis scales across multiple charts meant to be compared can make similar values look different or different values look deceptively similar.
- Dual-axis charts can visually suggest a correlation between two metrics that doesn't actually exist, depending on each axis's chosen relative scaling.
- 3D effects like a 3D pie chart distort perceived proportions, systematically making front-facing slices look larger than their true underlying value.