An unexplained highlight gets read as a fact about the data itself, not a designer's choice

What it is

Readers default to reading visually prominent elements as "special in the data itself" — the bolded line becomes "the most important series," the enlarged point becomes "an unusually large value." When the real basis for emphasis is "this series launched this quarter" or "this is the presenter's team," an unstated basis lets readers convert the designer's emphasis into a data property. Emphasis is a design judgement, not a data fact, and only words can police that boundary.

Why it happens

The misread comes from the automatic link between salience and importance: years of chart reading teach the heuristic "the more prominent the data, the more important it is," which serves well until it meets human-injected emphasis. The designer has inserted a dimension absent from the data ("who was selected") without supplying a legend or note for the new dimension, so readers convert it back into the most familiar interpretation — data importance. Stating the basis makes the new dimension explicit: one line — "bold marks projected values" — pulls readers off the misread track.

Where it stops holding

Two kinds of highlight may skip the note: interaction-generated temporary highlights (hover, selection) are self-explanatory — the user caused them; conventionalised highlights can be spared for mature audiences (bolding "our company" in an annual-report chart is an industry norm). Everything else must state its basis, analytic emphases above all ("highest growth," "below the safety line"), because they look the most like data properties. The note belongs beside the chart or the target — a sentence buried paragraphs into the body text still misses the first-glance misread.

Applying it

• Attach a note to every human-chosen highlight: "bold = our company," "orange = pilot cities," placed next to the chart or in the caption.
• For interactive highlights, prompt the basis once at first appearance and let convention carry it afterwards.
• Verification: ask readers to explain why one line differs from the rest; a data-based answer ("it is the largest") instead of the stated basis means the note was not read or not written — fix and retest.

Related

• Same group: U5.05.1 Highlighting works through contrast with the rest, not through its own intensity · U5.05.2 Dimming the rest disturbs the whole less than amplifying the target · U5.05.3 De-emphasised elements keep a readable minimum contrast · U5.05.4 Highlighting too many items at once is highlighting none
• Nearby: U4.06.5 Semantic colour needs a text label to be judged on its own · U10.03.3 Narrative wording often outruns what the data supports
• Search terms: highlight annotation · emphasis disclosure · salience misattribution