L2.2: Describing Categorical Data - Charts of Categorical Data

​

Learning outcomes

• identify common charts for categorical data
• distinguish bar chart from pie chart
• choose an appropriate visual for a categorical dataset
• interpret categorical charts correctly

​

Main chart types for categorical data

​

Bar chart

• Uses separate bars for each category.
• Height of each bar shows frequency or percentage.
• Bars are separated because categories are distinct.

• comparing category sizes clearly
• many categories
• when exact comparisons matter

​

Pie chart

• Shows categories as parts of a whole circle.
• Each sector angle represents proportion of the total.

• small number of categories
• showing part-to-whole relationships

​

Bar chart vs histogram

• Bar chart:
  • categorical data
  • bars separated
• Histogram:
  • numerical continuous data grouped into intervals
  • bars usually touch

• Do not use histogram rules for categorical data questions.

​

Interpreting charts

1. Which category is largest?
2. Which is smallest?
3. Are values shown as counts or percentages?
4. Is the graph based on one whole set?

​

Example

• bar chart using frequencies
• pie chart using percentages

​

Exam hints and traps

• Pie charts are less effective when many categories exist.
• Bar chart bars should have equal width.
• The y-axis label must be meaningful.
• A pie chart should represent one full total, not unrelated totals.

​

Quick practice

1. Which chart is better for comparing 8 categories quickly?
2. Which chart emphasizes “part of the whole”?
3. Why is histogram not suitable for blood group data?

​

Answer key

1. Bar chart
2. Pie chart
3. Blood group is categorical, not continuous grouped numerical data.