Learn
Interactive·

Data Visualization & Graph Reading

A good graph is not decoration; it shows the comparison that matters without distorting it.

The same numbers emphasize different messages when shown as bars, lines, or slices. Choosing a chart and deciding where an axis begins are not merely design choices—they can change how the result is interpreted.

Good visualization makes the relevant comparison easy and communicates magnitude, change, distribution, or relationship without unnecessary decoration.

Key question

Which graph, axis, and visual encoding best match the comparison you want readers to make?

1

Concept

What comparison does a graph invite?

A graph determines which comparison the reader can make easily

Analytical questionGraph to consider first
Compare categoriesBar chart
Show change over timeLine chart
Relate two numeric variablesScatterplot
Show the distribution of one variableHistogram or boxplot
Show share of a whole100% stacked bar or pie chart
Show geographic patternMap, distinguishing counts from rates
2

Why It Matters

Which graph should you choose?

Chart type should match both the data structure and the question

If time order is central, isolated bars may weaken the sense of movement. Conversely, connecting unrelated categories with a line can imply continuity and intermediate values that do not exist.

SituationChoice to avoid
Many categories with small share differencesa pie chart that requires difficult angle comparisons
Irregular time gapsa line chart drawn with equal spacing
Magnitude comparison is the goalpictograms whose area or volume exaggerates differences
Comparing group distributionsshowing only mean bars
3

How It Works

Why do axes and scales matter?

Axes, units, and denominators determine the visual impression

Bar length is interpreted relative to zero, so the value axis of a bar chart should generally start at zero. A line chart may use a narrower range to show small changes, but the truncation and full context should be made explicit.

CheckQuestion to ask
Axis rangeDoes it start at zero, or has part of the scale been removed?
UnitsAre people, thousands, millions, dollars, or percentages clear?
DenominatorIs the rate based on the full population or only respondents?
Time period and spacingAre the start, end, and intervals accurate?
Sample and sourceWho is included, and where did the data come from?
Color and legendAre emphasis and classification rules consistent?
4

Example

Which designs create misleading impressions?

Inspect the comparison structure before reacting to visual size

  • 3D bars and perspective distort height and area.
  • Dual axes can be adjusted to make unrelated trends look aligned.
  • Middle segments in stacked charts lack a common baseline and are hard to compare.
  • Raw counts may simply reflect population or exposure differences.
  • Averages alone can hide distribution, sample size, and uncertainty.

Reading order

Before accepting the headline, inspect the axes, units, denominator, period, sample, and source—then read the actual pattern in the data.

5

Interactive

Change the graph type and axis

Change the graph type and axis

Display the same values as bars and as a line and compare which relationship each emphasizes. Then truncate the vertical axis and see how the visual impression grows even though the numerical differences stay the same.

043856872747981
Starting at zero preserves bar-length proportions.

What to notice

A graph can create a strong impression without changing the underlying values. Check the comparison, axis, unit, and time period before interpreting the shape.

Key takeaways

  • Choose a graph that matches the analytical question and data structure.
  • Axis range, units, denominator, and time period can substantially change interpretation.
  • 3D effects, dual axes, and inappropriate area encodings can distort differences.
  • Inspect the comparison structure and data source before accepting the graph’s message.
Now build graphs from a real datasetOpen in Lab
Descriptive statistics curriculum complete