Market Research: Foundations to Practice · Communicating Insights and Research Careers
Visualizing and Presenting Data Honestly
A chart can technically display accurate numbers while still creating a visually misleading impression, and this happens more often through careless default settings than deliberate intent. This chapter covers the common failure modes.
A bar chart with a y-axis that doesn't start at zero can make a genuinely small difference between two values look dramatically larger than it actually is, a common and easily avoidable visual distortion that occurs more often through careless default chart settings than through any deliberate intent to mislead, but that still meaningfully misleads a viewer regardless of the chart creator's actual underlying intent either way.
- A bar chart with a truncated (non-zero) y-axis can make a small difference look dramatically larger, more often through careless defaults than deliberate intent.
- Choosing the right chart type for the data (line for trends, bar for categories, scatter for relationships) matters for honest communication as much as the numbers themselves.
- Showing margin of error or confidence intervals alongside point estimates helps viewers judge whether an apparent difference is real or normal sampling variation.
- Certain chart types like 3D pie charts are prone to misleading by default and are generally best avoided in professional research communication.