Completeness

Why I begin with completeness

When a chart is incomplete, it fails before interpretation even begins. If I cannot tell what is being measured, for whom, in which unit, or under what conditions, then the chart is already in trouble. That is why I start here.

In this chapter, we look at the rules that make a chart understandable before we even start debating whether it is elegant or persuasive.

The rules I group under completeness

• Rule 1. Clear title
• Rule 2. Axis labels
• Rule 3. Units and scale clarity
• Rule 4. Legend clarity
• Rule 5. Annotation context
• Rule 6. Uncertainty cues

Rule 1: Clear title

When I see a vague or generic title, I immediately know that the chart is asking the reader to do extra interpretive work. A good title should not merely name a topic. It should help frame what the chart is about.

• I ask students to specify what is measured, for whom, and over what period.
• Titles such as Sales or Performance are usually too weak to do real explanatory work.

Rule 2: Axis labels

Axes are not decoration. They tell the reader what the values mean. Without them, even a visually clean chart can remain ambiguous.

• Missing labels force the reader to guess.
• Generic labels such as Value are only acceptable when the context is already unmistakable.

Rule 3: Units and scale clarity

I insist on units because numbers by themselves are not enough. A value without a unit is incomplete information.

• I expect students to say whether a value is a percentage, a count, euros, kilograms, or something else.
• This matters even more when a figure circulates outside the notebook or slide deck in which it was created.

Rule 4: Legend clarity

Whenever a chart contains several series, the reader needs a clean bridge between the marks and their meaning. If that bridge is weak, reading slows down immediately.

• A missing legend delays interpretation.
• A vague legend weakens the chart even when the graphic itself is technically correct.

Rule 5: Annotation context

Some charts need more than labels and scales. They need a visible message. In class, I often tell students that if a chart contains one insight worth noticing, the chart should help the reader notice it.

• I encourage students to annotate an outlier, a turning point, or a key comparison.
• A short sentence inside the figure can sometimes do more than a paragraph outside it.

Rule 6: Uncertainty cues

When values come from estimates, samples, or models, uncertainty is often part of the information. Leaving it out can make a result appear more definitive than it really is.

• Confidence intervals, error bars, or uncertainty bands can help reduce overconfidence.
• I treat this rule as a strong teaching prompt, even when it is not always an automatic failure condition.