Quiz: Typography, Annotation, and the Words on Your Chart

20 questions. Aim for mastery (18+). If you score below 14, revisit the relevant sections before moving to Chapter 8.

Multiple Choice (10 questions)

1. The chapter defines a "self-explanatory chart" as one that:

(a) Is entirely code-generated without manual intervention (b) Can be read and understood without any surrounding text (c) Contains every possible data point in the underlying dataset (d) Uses a specific chart type from the chart selection matrix

Answer **(b)** Can be read and understood without any surrounding text. The self-explanatory standard means the chart carries its own context: a title that explains what it is about, axis labels with units, annotations for key features, and source attribution for trust. The reader can understand the chart without needing to consult an accompanying article, caption, or document. The chapter argues this standard matters because charts travel — they end up in slides, emails, and social media posts where the original context is lost.

2. Which of the following is the best example of an action title?

(a) "Quarterly Revenue, 2020-2024" (b) "Revenue Over Time" (c) "Meridian Corp Revenue Grew 18% in 2024, the Best Year Since 2019" (d) "Bar Chart of Quarterly Revenue"

Answer **(c)** "Meridian Corp Revenue Grew 18% in 2024, the Best Year Since 2019." An action title states the finding — what the reader should conclude — not just the topic. Options (a) and (b) are descriptive titles that name the chart's subject but do not tell the reader what to look for. Option (d) describes the chart type, which is even less useful. Only (c) is a declarative claim with specific numbers that tells the reader what the data is saying.

3. The chapter argues that the single most impactful change most practitioners can make is:

(a) Switching from matplotlib to a newer library (b) Using colorblind-safe palettes (c) Rewriting chart titles from descriptive to action titles (d) Adding interactive tooltips to static charts

Answer **(c)** Rewriting chart titles from descriptive to action titles. This is the chapter's threshold concept: "The title states the finding." The action title converts a chart from a puzzle the reader has to solve into a presentation with a clear message. The change takes under a minute per chart and is reversible, and it produces a dramatic improvement in reader comprehension in the 5-second reading window.

4. Which of the following is NOT one of the five typographic principles introduced in Section 7.2?

(a) Use a single, legible, sans-serif font family (b) Establish a size hierarchy between title, labels, and tick labels (c) Use color (not weight) for emphasis (d) Leave whitespace around text

Answer **(c)** Use color (not weight) for emphasis. The chapter explicitly says the opposite: use **weight** (bold, semi-bold, regular) for emphasis, not color. Color is precious in data visualization because it carries encoding information for the data itself; spending color on text emphasis means less color available for the data. Weight is a free emphasis channel. The other three options are direct paraphrases of principles from Section 7.2.

5. The chapter recommends that titles and subtitles should be:

(a) Centered above the plotting area (b) Left-aligned with the plotting area's left edge (c) Right-aligned with the plotting area's right edge (d) Rotated 90 degrees to run up the left margin

Answer **(b)** Left-aligned with the plotting area's left edge. Left-aligned titles create a consistent visual edge with the y-axis and the leftmost data, which anchors the chart visually. Centered titles float above the plotting area with no visual relationship to it, which looks less professional. This alignment choice is a small detail, but it is one of the details that separates thoughtful chart design from default-chart design.

6. An annotation on a chart is best described as:

(a) A decorative graphic element that adds visual interest (b) A small piece of text placed directly on the data that explains a specific feature (c) A formal caption below the chart explaining the methodology (d) A legend that identifies multiple series by color

Answer **(b)** A small piece of text placed directly on the data that explains a specific feature. Annotations are specific, short, and attached to particular features of the data — an outlier, an inflection point, a contextual event, a threshold. They are the text that does the most work per word because they carry information about a precise location in the data. Captions (option c) and legends (option d) are separate chart text categories with different roles.

7. The chapter recommends one to three annotations per chart because:

(a) More than three annotations violates most style guides (b) Annotation functions in matplotlib have a three-annotation limit (c) Annotations compete for attention, and more than three diffuse the reader's focus (d) Three is the smallest number that is statistically significant

Answer **(c)** Annotations compete for attention, and more than three diffuse the reader's focus. The annotation budget is about focus, not arbitrary limits. One annotation creates one prominent feature. Three annotations create three points of focus. Ten annotations mean the reader cannot tell which ones matter most, and the chart becomes a wall of text with no narrative. When you find yourself wanting more than three annotations, the chart is usually trying to tell multiple stories at once, and the right response is often to split it into multiple charts.

8. The chapter argues that direct labeling is usually preferable to a legend because:

(a) Legends are forbidden by most corporate style guides (b) Direct labels eliminate the eye movement between the data and a separate legend box (c) Direct labels use less ink than legend boxes (d) Direct labels are easier to implement in matplotlib

Answer **(b)** Direct labels eliminate the eye movement between the data and a separate legend box. When a legend is separate from the data, the reader's eye must move between the data and the legend repeatedly, and in a dense chart the reader can lose track of which line is which. Direct labels place the series name next to the series itself, so identification happens without any eye movement. This is especially valuable in the 5-second reading window, where any extra eye movement is a cost.

9. Which of the following is the correct recommendation for tick density according to the chapter?

(a) The more tick marks, the more precise the chart (b) Use as many ticks as the default library produces (c) Use fewer ticks than the default — typically five or six major ticks per axis is enough (d) Use ticks at every data point

Answer **(c)** Use fewer ticks than the default — typically five or six major ticks per axis is enough. Default plotting libraries often produce more ticks than the reader needs. Fewer ticks means less visual noise, cleaner tick labels, and more room for the data. The reader can interpolate between ticks without difficulty. When in doubt, cut the tick count in half and see if the chart still reads correctly.

10. The chapter's position on on-image source attribution is that:

(a) It is optional for exploratory charts but required for publications (b) It is non-negotiable and should be built into your default workflow for every published chart (c) It is a citation convention that matters only for academic papers (d) It should appear only in the caption, never on the chart itself

Answer **(b)** It is non-negotiable and should be built into your default workflow for every published chart. The chapter's argument is that source attribution is a trust mechanism, not just a citation convention, and that it must be on the chart itself (not in a separate caption) because captions get stripped when charts are shared. The discipline is cheap — one line of text — and the payoff is a consistent baseline of trust across all of your work.

True / False (5 questions)

11. "An action title is editorializing because it states a finding, and editorializing is always wrong in data visualization."

Answer **False.** Chapter 4 established that every chart is an editorial — there is no neutral chart. A descriptive title does not make a chart neutral; it only hides the editorial stance. An action title makes the stance explicit, which is arguably the more honest approach. The ethical line is not between action titles and descriptive titles but between titles that match the evidence and titles that overstate it. A defensible action title ("Revenue Grew 18%" when the data shows 18% growth) is not "editorializing" in any objectionable sense.

12. "A chart that has been carefully decluttered (per Chapter 6) is automatically self-explanatory and does not need titles, annotations, or source attributions."

Answer **False.** Decluttering and becoming self-explanatory are different operations. Decluttering removes visual noise so the reader can find the data. Becoming self-explanatory requires adding the right words so the reader knows what the data means. The two operations are both necessary. A clean but wordless chart can be read (the eye finds the shapes) but cannot be understood (the reader does not know what they are looking at or what to conclude).

13. "Sans-serif fonts are generally preferable to serif fonts for chart text, especially on screen."

Answer **True.** The chapter recommends sans-serif fonts (Helvetica, Arial, Inter, Source Sans Pro, IBM Plex Sans) for data visualization. Serifs are visual features that help the eye follow long lines of body text in books and newspapers, but they become noise at the small sizes used in chart labels, especially on screens. For charts, sans-serif is cleaner and more legible.

14. "For a chart with two series, using a legend is always preferable to direct labeling because the reader's working memory can easily handle two categories."

Answer **False.** With two series, direct labels are usually cleaner and cheaper in reader effort than a legend, even though the working memory cost is manageable in either case. Direct labels avoid the eye movement to and from the legend box and produce a visually cleaner chart. Legends are appropriate for large numbers of series, for crossing lines, and for interactive contexts with hover; they are not the default for two-series charts.

15. "A chart that includes a source attribution in its caption but not on the image itself is following the chapter's attribution principle."

Answer **False.** The chapter explicitly argues that attribution must be on the chart itself, not in a separate caption, because captions get stripped when charts are shared, screenshotted, or embedded. The attribution has to travel with the chart, which means it has to be inside the image. A chart with attribution only in a caption will be shared in a form that removes the attribution, defeating the purpose of citing the source.

Short Answer (3 questions)

16. In three to four sentences, explain the difference between a "descriptive title" and an "action title," and state when each is appropriate.

Answer A **descriptive title** states what the chart is about — the topic. Examples: "Quarterly Revenue" or "Global Temperatures Over Time." An **action title** states the finding — what the reader should conclude. Examples: "Revenue Grew 18% in 2024" or "Global Temperatures Have Risen 1.2 Degrees Since 1900." Descriptive titles are appropriate for exploratory charts, reference tables, and academic publications with neutral conventions. Action titles are appropriate for business communication, data journalism, dashboards, and any context where the reader needs to grasp the main message quickly.

17. The chapter describes the "5-second test" for evaluating chart text. Explain the test and describe what the test reveals that is not obvious from looking at the chart yourself.

Answer The 5-second test asks a first-time viewer to describe a chart in their own words after seeing it for exactly five seconds. A chart passes if the viewer can state (1) what the chart is about, (2) what the main finding is, and (3) what the units are. The test reveals gaps in self-explanation that the chart maker cannot see because they already know the data. The maker's prior knowledge lets them see the message even when the chart does not provide it; a first-time viewer cannot, and the test surfaces the missing context (usually a missing action title, missing units, or missing annotation).

18. List three specific tick-label formatting improvements recommended in Section 7.5 and explain why each one improves readability.

Answer (1) **Use thousands separators** (5,000,000 instead of 5000000) — makes large numbers immediately scannable without counting digits. (2) **Scale to natural units and state the scaling in the axis label** (5M with "Revenue in millions" instead of 5000000) — puts the values in the form the reader naturally thinks in. (3) **Use the minimum number of decimal places needed** (5, 6, 7, 8 instead of 5.00, 6.00, 7.00, 8.00) — removes visual noise from unnecessary digits. Also acceptable: format dates for human reading (Jan 2020 instead of 2020-01-01), reduce tick density (five or six major ticks instead of ten or more), right-align numerical tick labels on the y-axis so digits line up.

Applied Scenarios (2 questions)

19. You have produced a line chart of Meridian Corp's quarterly revenue for the past three years. The chart currently has a default title ("Chart1"), no subtitle, y-axis labeled "Revenue" with tick marks at 4000000, 4200000, 4400000, 4600000, 4800000, 5000000, 5200000, 5400000. There is no source attribution. The chart is for a quarterly business review slide deck. Identify every problem with the chart text and prescribe a complete fix.

Answer **Problems identified:** - Title is a default stub ("Chart1") with no descriptive or action content. - No subtitle providing context (time range, scope). - Y-axis label "Revenue" lacks units — reader does not know if values are USD, EUR, etc. - Tick labels are raw numbers without thousands separators or scaling, which is hard to read. - Too many ticks (nine) — too dense. - No source attribution. **Prescription:** - **Title (action):** "Meridian Corp Revenue Grew 18% in 2024, Reaching $5.4B in Q4" - **Subtitle:** "Quarterly revenue, Q1 2022 – Q4 2024" - **Y-axis label:** "Revenue (USD millions)" - **Tick labels:** Scale to millions, use five ticks at 4000, 4400, 4800, 5200, 5600 (or similar). Display as "4,000", "4,400", "4,800", "5,200", "5,600" or use "4.0M", "4.4M", "4.8M", "5.2M", "5.6M" — either works. - **Source attribution (bottom, 8pt, muted):** "Source: Meridian Corp Finance Department, Q4 2024 earnings report. Internal data." - Also consider: an annotation on the Q4 2024 bar stating "+18% YoY" to reinforce the action title.

20. You are designing a chart of global COVID-19 case rates across 10 countries for a general public audience. The chart will appear on a news website and will likely be shared on social media. You are deciding between (a) a line chart with a color legend in the upper right, and (b) a line chart with direct labels on the ends of the lines and the two highlighted countries (the reader's country and a comparison country) in bold colors, with the other eight countries drawn as muted gray lines.

(a) Which design do you choose and why? (b) Write an action title for the chart. (c) Identify one additional piece of chart text you would include that the default design would omit.

Answer **(a) Design (b) is the better choice.** The direct labels eliminate the eye movement between the data and a legend box, which matters in the 5-second glance that social media readers will give the chart. The highlight strategy (two colored lines against eight muted gray) focuses attention on the comparison that matters to the reader's country while preserving context from the other countries. Design (a) with ten-color legend would force the reader to match each line to its legend entry, which is slow and error-prone. **(b) Example action title:** "The United States Had Higher Case Rates Than Most OECD Countries Through Most of 2022" — or any similar specific claim supported by the data. A descriptive title like "COVID-19 Case Rates by Country" would be weaker because it does not tell the reader what the chart is saying. **(c) Additional chart text:** A **subtitle** stating the time range, data source, and unit of measurement — e.g., "New cases per 100,000 people, 7-day rolling average, Jan 2020 – Dec 2022. Source: Our World in Data." Also strongly recommended: an **annotation** on the peak of each highlighted country, stating the peak value and the date. Either addition would significantly improve the chart's readability for a first-time viewer.

Review your results against the mastery thresholds at the top. If you scored below 14, revisit Sections 7.1 (self-explanatory standard) and 7.3 (action titles) — those are the foundational concepts for the rest of the chapter. Chapter 8 builds on the typographic and annotation skills from this chapter to address layout and composition of multiple charts on a single figure.