Track 3 · Visualising and Exploring Data · Lesson 2 of 11
Bar charts for categories
Draw and read bar charts of categorical data — sorted, grouped or stacked — and report proportions with the right denominator.
A retail chain wants to know how customers pay, and whether loyalty-card members pay differently. A marketing team wants to know which channel “performs best”. Both answers come from bar charts of categories — and both are easy to get wrong by sorting badly, stacking the wrong way or dividing by the wrong total.
A bar chart shows one bar per category, with the bar's length proportional to a count, a percentage or another amount. Three rules make bar charts honest and quick to read:
- Start the value axis at zero. The eye compares lengths; a bar cut off at 50 makes 60 look twice as big as 55.
- Sort nominal categories by size (largest first) so the ranking can be read immediately. Keep ordinal categories (sizes, ratings, age bands, job levels) in their natural order.
- Use horizontal bars when labels are long, so they can be read without tilting your head.
Included with a Formula Quest membership
It has 8 graded exercises with feedback, hints and worked solutions. After it you will be able to:
- Draw bar charts from a zero baseline, sorted for nominal categories and in natural order for ordinal ones
- Choose between grouped, stacked and 100% stacked bars according to the question
- Compute row and column percentages and state the denominator
- Distinguish volume (counts) from efficiency (rates) before declaring a winner