Data Visualization Basics

A chart's job is to make a pattern in the data obvious at a glance — and picking the wrong chart type, or drawing the right one badly, can just as easily hide a pattern or invent one that isn't there.

Match the chart to the question

A few reliable defaults cover most situations: a bar chart compares a value across separate categories (sales by region). A line chart shows how something changes over a continuous range, usually time (revenue by month). A scatter plot shows the relationship between two numeric variables (advertising spend vs. sales). A histogram shows the distribution of a single variable (how ages are spread across a customer base). Reaching for the wrong one — a line chart for unordered categories, say — tends to imply a trend or connection that doesn't actually exist.

A truncated axis exaggerates everything

Starting a bar or line chart's y-axis somewhere other than zero is one of the most common ways a chart misleads, even accidentally. A change from 51% to 53% looks like a tiny move on an axis running 0-100%, but looks like a dramatic surge on an axis running 50-54% — same data, wildly different impression.

Too much information kills the point

A chart with a dozen overlapping lines, six colors, and three different scales usually communicates less than a chart with one clear line and a title stating the takeaway. If a chart needs a long paragraph of explanation to be understood, it's usually a sign to simplify the chart rather than lengthen the explanation.

3D and pie charts, used carefully

3D effects on a bar or pie chart usually distort the actual proportions being shown, purely for visual flair — avoid them. Pie charts themselves work fine for a small number of categories that clearly sum to a whole, but become hard to read accurately once there are more than four or five slices, since the human eye is genuinely bad at comparing angles.