Selecting & Interpreting Data Representations

Line Graphs are superior for showing changes over time (time-series data) because the slope of the line indicates the rate of change.

Box Plots provide a high-level summary of spread and outliers but do not show the frequency of specific values.

Scale Integrity: Always check if the vertical axis starts at zero. A 'broken' or truncated axis can exaggerate small differences between data points.

Visual Clarity: Elements like excessively thick lines or overly bright colors can distract the viewer or make it difficult to read precise values from the axes.

The 3D Distortion Problem: 3D charts (especially 3D pie charts) are often misleading because perspective causes segments at the 'front' to appear larger than segments at the 'back', even if their actual values are smaller.

Justification is Key: When asked to choose a representation, always justify your choice using three pillars: the data type (e.g., continuous), the audience (e.g., expert), and the goal (e.g., showing spread).

Check for Keys: If a diagram uses colors or symbols without a legend/key, it is technically incomplete and cannot be fully interpreted.

Compare Proportions vs. Totals: Be careful when comparing two pie charts of different sizes; they show proportions, but without knowing the total frequency ($n$) for each, you cannot compare the actual number of items in each category.

Sanity Check: Always look at the total frequency. If a bar chart shows 50 people and a pie chart shows 25%, ensure the calculation $0.25 \times 50 = 12.5$ makes sense in context (e.g., you can't have half a person).