Assignment_1

Introduction

This report evaluates two effective and two ineffective/misleading figures from external sources. The analysis uses principles from Gelman & Unwin and Cleveland & McGill—particularly perceptual accuracy (position/length vs area/angle), integrity of scales and baselines, clarity of labeling, and avoidance of non‑informative decoration. Each figure is shown with its source and discussed in context.

Effective Figures

Figure 1 — Minard’s “Carte Figurative” of Napoleon’s 1812–1813 Campaign [1]

Source: Charles Joseph Minard (1869). Link: https://en.wikipedia.org/wiki/Charles_Joseph_Minard

Context: Minard’s “Carte Figurative” is often considered one of the most effective visualizations ever created because it communicates a complex historical event with remarkable clarity. The width of the flow encodes the size of Napoleon’s army, and as the path progresses geographically, the viewer can immediately see how the army’s numbers dwindle over time. The use of color differentiates the advance (beige) from the retreat (black), and the lower panel adds another dimension by showing the rapidly falling temperatures during the retreat.

By combining geography, time, direction, temperature, and troop size into a single coherent display, the figure tells a powerful story without relying on decorative elements. The encodings rely on position and length, which Cleveland and McGill identify as highly accurate perceptual channels, rather than on area or 3D distortion. In line with Gelman and Unwin’s principles, the graphic minimizes chartjunk and maximizes data-ink ratio, resulting in a figure that is both visually striking and analytically rigorous.

Figure 2 — John Snow’s 1854 Cholera Map (Broad Street Outbreak) [2]

Snow’s map showing cholera deaths clustered around the Broad Street pump.

Source: John Snow (1855).(Public domain reproduction). Link: https://en.wikipedia.org/wiki/1854_Broad_Street_cholera_outbreak#/media/File:Snow-cholera-map-1.jpg

Context: John Snow’s cholera map is effective because it makes the source of a deadly outbreak immediately visible. Each bar marks the location of cholera deaths, and when plotted on the neighborhood street map, the cases clearly cluster around the Broad Street water pump. The visual design is straightforward but powerful: by simply placing cases where they occurred, the map uncovers a pattern that might have been missed in raw tables or text descriptions.

The figure also succeeds because it provides the right amount of context. The street layout, pump locations, and death counts are included, but there is no unnecessary decoration to distract from the data. This balance allows readers whether experts or ordinary citizens to connect geography to public health outcomes. The clarity of the spatial encoding made it possible for Snow to argue convincingly that cholera was waterborne, showing how a well-designed visualization can literally change the course of science and save lives.

Ineffective / Misleading Figures

Figure 3 — USA Today “Common Injuries Children Suffer” [3]

USA Today Snapshot using crutch steps to encode percentages; lengths/areas are not proportional to values.

Source: USA Today Snapshot by Shannon Reilly and Frank Pompa. Data: American Academy of Orthopaedic Surgeons. Link: https://www.statisticshowto.com/probability-and-statistics/descriptive-statistics/misleading-graphs/

Context:
At first glance, this USA Today Snapshot seems engaging because it uses the image of crutches to show injury statistics, but the design actually misleads the viewer. The “steps” of the crutch are drawn at different sizes that do not match the true percentages. For instance, the difference between 21.7% and 21.5% is visually exaggerated even though it is practically negligible. By stretching and shrinking the rungs unevenly, the figure distorts how readers perceive the numbers.

The problem is made worse by the heavy use of decoration. The crutch illustration draws attention away from the data, yet it fails to provide a clear axis or consistent scale that would allow the reader to check the proportions accurately. Instead of helping us understand which injuries are most common, the graphic distracts with a metaphor that adds confusion. This violates the principle of avoiding “chartjunk” and shows how poor design choices can make a simple dataset look more dramatic than it really is.

Figure 4 — Newspaper Sales Bars with Truncated Baseline [4]

Bar chart comparing newspaper sales where Y-axis starts near 420k instead of 0, exaggerating differences.

Source: Newspaper graphic (The Times vs Daily Telegraph full‑price sales). Link: https://flowingdata.com/2011/12/12/fox-news-still-makes-awesome-charts/ Context:
This Fox News unemployment chart is misleading because the Y-axis does not start at zero. Instead, it begins around 8%, which makes small changes in the unemployment rate look far more dramatic than they really are. For example, the decline from 9.0% to 8.6% appears as if unemployment has been cut in half, when in reality the difference is less than half a percentage point. By presenting the data in this way, the graphic exaggerates the trend and risks misleading viewers who expect bar lengths to represent values from a common baseline.

A more honest design would either start the Y-axis at zero or use a different chart type, such as a simple line plot with clearly labeled values. That approach would highlight the modest decline without distorting the viewer’s perception. According to Cleveland and McGill, position and length should be used for accurate comparisons, but only when they are anchored to a true baseline. This chart fails that test, and in doing so undermines the integrity of the data it is supposed to show.

Discussion

Looking across these four figures, several clear lessons emerge that connect directly to the principles outlined by Gelman & Unwin and by Cleveland & McGill. The effective figures—Minard’s map of Napoleon’s march and John Snow’s cholera map—both rely on position and length as their primary visual encodings. Cleveland and McGill show that these channels are among the most accurate for human perception, which explains why both visualizations succeed in communicating complex stories with clarity. In addition, they avoid unnecessary decoration, keeping the focus on the data itself. This matches Gelman and Unwin’s emphasis on maximizing the data-to-ink ratio and ensuring that graphical choices serve the message rather than distract from it.

By contrast, the misleading figures show what happens when these principles are ignored. The USA Today injury graphic uses a pictorial metaphor that introduces distortions, turning a simple frequency comparison into a confusing and exaggerated display. The Fox News unemployment chart manipulates the Y-axis baseline, creating the illusion of a dramatic change where only a small difference exists. Both examples demonstrate how poor design choices can undermine trust and mislead audiences, even when the underlying data is accurate.

Together, these cases highlight the importance of design integrity in data visualization. Effective graphics respect perceptual limits, use honest scales, and provide sufficient context for interpretation. Ineffective ones do the opposite, whether by exaggerating small differences or cluttering the display with unnecessary imagery. As the readings stress, good visualization is not just about making charts attractive—it is about making data truthful, clear, and useful for informed decision-making.

References

1. Minard, C. J. (1869). Carte figurative des pertes successives en hommes de l’armée française dans la campagne de Russie 1812–1813. Public domain. Retrieved from https://en.wikipedia.org/wiki/Charles_Joseph_Minard
   
2. Snow, J. (1855). On the Mode of Communication of Cholera. London: John Churchill. Retrieved from https://en.wikipedia.org/wiki/1854_Broad_Street_cholera_outbreak
   
3. VizWiz. (2010). “USA Today Visualization.” Retrieved from https://www.statisticshowto.com/probability-and-statistics/descriptive-statistics/misleading-graphs/
4. FlowingData. (2011). “Fox News Still Makes Awesome Charts.” Retrieved from https://flowingdata.com/2011/12/12/fox-news-still-makes-awesome-charts/