Choosing the Right Chart

Lesson 2 of 8 · Data Visualization Fundamentals · Beginner

Choosing the right chart type is one of the most important decisions in data visualization. The wrong chart can obscure the very pattern you are trying to show. The right chart makes the answer obvious.

Start with the Question

Chart selection follows from the question, not from the data format. Ask: what relationship, comparison, or pattern am I trying to show? The answer points toward a chart category.

Common Chart Categories and When to Use Them

Comparison across categories — Bar chart

Bar charts compare values across distinct categories. A horizontal bar chart works well when category names are long or when there are many categories. A vertical bar chart (column chart) works well for a smaller number of categories with short labels.

Example: comparing average monthly visits across five community programs.

Change over time — Line chart

Line charts show how a value changes across an ordered sequence, most commonly time. The connected line implies continuity between points. Use a line chart when the data points represent a continuous or regularly measured series.

Example: monthly service requests over two years.

Relationship between two numeric variables — Scatter plot

Scatter plots show whether two numeric variables tend to move together. Each point represents one observation. Scatter plots reveal correlation, clusters, and outliers. They do not prove causation.

Example: plotting program attendance against facility capacity to see whether larger facilities attract more visitors.

Distribution of a single numeric variable — Histogram

Histograms show how values are distributed across a range by grouping them into intervals (bins). They are not bar charts — the bins are continuous, and the chart shows frequency or count within each range.

Example: distribution of response times for maintenance requests, grouped into 0–1 day, 1–3 days, 3–7 days, and so on.

Part-to-whole — Use with caution

Pie charts and donut charts show how parts contribute to a whole. They work only when there are few categories (typically two to four), the differences are large enough to see, and the exact values are labelled. With many categories or similar values, pie charts become unreadable. A stacked bar chart or a simple table often communicates part-to-whole relationships more clearly.

Location matters — Map

Maps are appropriate when geographic location is part of the question. If you are asking "where are service gaps?" or "which neighbourhoods have the highest usage?", a map may help. If location is not part of the question, a map adds complexity without adding insight.

Precise lookup — Table

When readers need to find exact values, compare specific numbers, or reference individual records, a well-formatted table is often clearer than any chart.

Things to Avoid

• Three-dimensional effects — 3D bars, 3D pies, and perspective effects distort values and make comparison inaccurate.
• Pie charts with many slices — more than four or five categories become impossible to compare visually.
• Line charts for unordered categories — a line implies a sequence; connecting unordered categories implies a relationship that does not exist.
• Maps when location is irrelevant — geographic display does not add insight when the question is not spatial.

Practical Example

A parks department wants to show three things: (1) which parks received the most maintenance requests last year, (2) how total requests changed month by month, and (3) what proportion of requests were completed within target time. The first question calls for a bar chart. The second calls for a line chart. The third — with only two categories (within target, outside target) — could use a simple bar chart or a single sentence with a percentage.

Common Mistakes

• Using a line chart for unordered categories.
• Using a pie chart with more than four or five slices.
• Adding 3D effects that distort values.
• Choosing a chart type because it looks impressive rather than because it answers the question.
• Using a map when the question is not geographic.

Key Takeaways

• Chart selection follows from the question, not from the data format.
• Bar charts compare categories; line charts show ordered trends; scatter plots show relationships; histograms show distributions.
• Pie charts work only with few categories and large differences.
• Maps are appropriate when location is part of the question.
• Tables remain useful when precise lookup matters.
• Avoid 3D effects, novelty formats, and charts inappropriate to the data type.