Skip to content

Chapter 6 Quiz

← Back to chapter

1. What are the four components of Edward Tufte's 4S Model for evaluating visualizations?

  1. Speed, Style, Substance, Scalability
  2. Significant, Supportive, Simple, Singular
  3. Statistical, Structural, Spatial, Sequential
  4. Sharp, Sleek, Succinct, Symbolic
Answer

B. Significant, Supportive, Simple, Singular The Tufte 4S Model requires that every good data graphic satisfy four criteria: Significant (shows data that matters), Supportive (reinforces the message), Simple (no unnecessary elements), and Singular (communicates one main idea). This serves as a checklist for evaluating any visualization before presenting it.

Concept tested: Tufte 4S Model

Bloom's level: Remember

2. What is the data-ink ratio, and what should communicators do with it?

  1. The ratio of colored ink to black ink; keep it balanced at 50/50
  2. The proportion of ink representing data versus decoration; maximize it
  3. The ratio of text to images on a slide; minimize text
  4. The number of data points per chart; keep it under ten
Answer

B. The proportion of ink representing data versus decoration; maximize it The data-ink ratio is Tufte's most quantifiable principle: the proportion of a graphic's ink (or pixels) used to represent actual data, as opposed to decoration, borders, backgrounds, and other non-data elements. The goal is to maximize this ratio by eliminating chartjunk -- visual elements that do not convey information.

Concept tested: Data-Ink Ratio

Bloom's level: Understand

3. Which chart type is most appropriate for showing the relationship between two continuous variables such as study hours and exam scores?

  1. Bar chart
  2. Pie chart
  3. Line chart
  4. Scatter plot
Answer

D. Scatter plot Scatter plots display data as individual points on a two-dimensional plane, with one variable on each axis. They are uniquely powerful for revealing relationships, correlations, and clusters between two continuous variables. Bar charts are for comparing categories, pie charts are for part-to-whole with few categories, and line charts are for trends over time.

Concept tested: Scatter Plots, Chart Selection

Bloom's level: Apply

4. Why does the chapter recommend against using pie charts in most situations?

  1. Pie charts use too much ink and violate the data-ink ratio principle
  2. Pie charts cannot display more than three categories
  3. Humans are poor at comparing angles and areas, making pie charts perceptually inaccurate
  4. Pie charts are considered outdated and unprofessional
Answer

C. Humans are poor at comparing angles and areas, making pie charts perceptually inaccurate Studies in perceptual psychology consistently show that people can compare bar lengths far more accurately than pie slice sizes. A pie chart with five slices of similar size becomes nearly impossible to rank visually, and adding 3D effects makes the problem worse. Pie charts are acceptable only with 2-3 categories where one dominates, with each slice labeled with its exact percentage.

Concept tested: Pie Chart Limitations

Bloom's level: Understand

5. A bar chart showing values of 98, 99, and 100 uses a y-axis that starts at 97. What form of visual deception does this represent?

  1. Cherry-picked time range
  2. Area and volume distortion
  3. Truncated y-axis (scale manipulation)
  4. Dual axes with mismatched scales
Answer

C. Truncated y-axis (scale manipulation) A truncated y-axis starts at a value other than zero, which exaggerates small differences. In this case, a 1% actual difference between values would visually appear enormous because the bars' heights are disproportionately different. Bar charts should almost always start their y-axis at zero because the audience compares bar lengths -- a bar that appears twice as tall should represent a value that is twice as large.

Concept tested: Scale Manipulation, Misleading Graphs

Bloom's level: Apply

6. When should you use a table instead of a chart to present data?

  1. When the dataset is very large and you need to show trends
  2. When the audience needs exact values and will look up individual data points
  3. When you want the audience to see patterns and relationships quickly
  4. When comparing two variables visually
Answer

B. When the audience needs exact values and will look up individual data points Tables are preferable when the audience's first question will be "What is the exact number for X?" Charts are better when the question is "What is the trend?" or "How does X compare to Y?" A useful approach is to use both -- a chart for the pattern and a table in an appendix for precise reference.

Concept tested: Tables vs Charts

Bloom's level: Understand

7. Which color palette type should be used when data ranges from negative to positive values with a meaningful midpoint?

  1. Sequential palette
  2. Diverging palette
  3. Categorical palette
  4. Highlight palette
Answer

B. Diverging palette Diverging palettes use two hues that diverge from a neutral center and are appropriate when data has a meaningful midpoint -- such as profit/loss (red through white to green) or below/above average. Sequential palettes are for low-to-high ranges. Categorical palettes distinguish unordered groups. Highlight palettes emphasize one element against a neutral background.

Concept tested: Color in Visualization

Bloom's level: Apply

  1. 1-2
  2. 4-8
  3. 10-15
  4. No limit, as long as they are well organized
Answer

B. 4-8 Effective dashboard design limits visualizations to 4-8 per screen. More than that overwhelms the viewer and violates the Simplicity Principle. If more metrics are needed, the chapter recommends creating multiple dashboard pages or tabs. The most important metric should be placed in the top-left corner (where Western readers look first) and given the largest chart.

Concept tested: Dashboard Design

Bloom's level: Remember

9. You are creating a chart for a presentation showing your team's API response times over the past 12 months, with an SLA target of 200ms. The response times improved from 450ms to 150ms during that period. Describe how you would design this visualization using principles from the chapter, including chart type selection, annotation strategy, and how to avoid common mistakes.

Answer

Chart type: A line chart is the correct choice because the x-axis represents time (12 months) and the goal is to show a trend -- the improvement in response times. Line charts connect data points to show continuity and progression, which is exactly what this data requires.

Annotation strategy: Add a horizontal reference line at 200ms labeled "SLA Target" to provide context -- this transforms the chart from "here are some numbers" to "here is how we compare to our goal." Annotate the starting point ("Jan: 450ms -- 2.25x above SLA") and the point where response times dropped below the SLA target ("Aug: 195ms -- first month below SLA"). Annotate the final point ("Dec: 150ms -- 25% below SLA"). Keep annotations to 2-3 key moments to avoid clutter.

Avoiding common mistakes: Start the y-axis at zero or at a reasonable baseline to avoid exaggerating the improvement through scale manipulation. Use a single color for the line with the SLA reference in a contrasting color (e.g., gray line for data, red dashed line for SLA). Do not add 3D effects, gradient fills, or decorative elements -- maintain a high data-ink ratio. Give the chart a takeaway title like "API response times dropped 67% and now exceed our SLA target" rather than a topic title like "API Response Times." Apply the 4S Model: the data is Significant (performance improvement), Supportive (reinforces the success narrative), Simple (one line with one reference), and Singular (one clear message about improvement).

Concept tested: Chart Selection, Annotation in Charts, Data-Ink Ratio, Tufte 4S Model

Bloom's level: Create

10. A colleague presents a chart that plots the company's revenue growth and the number of employees on dual y-axes, showing the two lines moving together. They claim this proves that hiring more people directly causes revenue growth. Using concepts from this chapter, evaluate this claim and explain what is misleading about the visualization.

Answer

This visualization is misleading in several ways identified in the chapter:

Dual axes with mismatched scales: Plotting two variables on different y-axes is one of the most common forms of visual deception. By adjusting the scales independently, you can make any two lines appear to move together -- or apart. The visual impression of correlation is an artifact of the axis scaling choices, not a property of the data.

Correlation vs. causation: Even if the two variables genuinely correlate, the visualization cannot prove that hiring causes revenue growth. Both might be driven by a third factor (e.g., market growth, new product launches), or the causation could run in the opposite direction (revenue growth enables hiring, not the other way around). The chart format implies a causal relationship that the data alone cannot establish.

Missing context: The chart likely uses a cherry-picked time range that shows both metrics rising. A longer time range might reveal periods where headcount grew but revenue did not, or vice versa.

Better approach: To honestly explore the relationship, use a scatter plot with headcount on one axis and revenue on the other, showing each data point as a dot. This would reveal whether a genuine correlation exists without the misleading dual-axis trick. If presenting both trends over time, use two separate charts with clearly labeled axes so the viewer can evaluate each trend independently. Add annotations noting that correlation does not imply causation.

Concept tested: Misleading Graphs, Scale Manipulation

Bloom's level: Evaluate