Displaying And Describing Categorical Data Distributions
Theory
Categorical data is summarised with frequency tables and bar charts. Convert raw counts to percentages using
For categorical data, the goal is to summarise how often each category occurs. The two main displays are frequency tables and bar charts.
A frequency table lists each category and how many times it occurs. A percentage column is often added:
A bar chart shows categories on the horizontal axis and frequency (or percentage) on the vertical axis. The bars are separated by gaps to indicate the categories are distinct, not continuous.
The mode is the most common category โ the one with the tallest bar (or the largest frequency in the table).
A segmented (stacked) bar chart is a single bar split into coloured sections, with each section representing a category's share. It's useful for percentage breakdowns or for comparing two groups side-by-side.
The first diagram is the favourite snacks bar chart from the worked example, with the mode (Chips) highlighted. The second compares a bar chart (categorical, with gaps) to a histogram (numerical, no gaps).
The only real formula is the percentage conversion.
Percentage from frequency
Frequency from percentage
Quick reference
| Quantity | How to find it |
|---|---|
| Mode | category with the largest frequency |
| Missing frequency | total minus the sum of the others |
| Combined % | add the relevant percentages, or add their frequencies then รท total ร 100% |
Building a frequency table
- List every category in the data โ including any with zero count.
- Count (tally) how many times each category appears.
- Add a percentage column using the formula above.
- Check that all frequencies add to the total and all percentages add to
(allowing small rounding error).
Drawing a bar chart
- Put categories on the horizontal axis, frequency (or percentage) on the vertical.
- Make all bars the same width, separated by equal-sized gaps.
- Start the vertical axis at
(no truncation). - Label both axes and give the chart a title.
Describing a categorical distribution
- State the mode โ the most common category.
- State the least popular category.
- Describe the general pattern: "one category dominates", "frequencies are roughly even", or similar.
Convert the percentage to a decimal and multiply by the total.
| Number | ||
Answer:
Use
Answer: about
The total of all categories equals
| Cat | ||
Answer:
Either add the individual percentages, or add the frequencies then divide by the total.
| Combined | ||
Check directly:
Answer:
Common pitfalls
Frequently asked questions
What is a frequency table?
A frequency table lists each category in one column and how many times it occurs in another. A percentage column is often added: percentage equals frequency divided by total, times 100.
What is the difference between a bar chart and a histogram?
Bar charts are for categorical data and have GAPS between the bars to show categories are distinct. Histograms are for numerical data and have bars TOUCHING because values run continuously into each other.
What is the mode of a categorical dataset?
The mode is the most common category โ the one with the tallest bar or largest frequency. Note that the mode is the CATEGORY itself (like 'Chips') and not the frequency number (like 36).
What is a segmented bar chart?
A segmented (stacked) bar chart is a single bar divided into coloured sections, with each section's length showing one category's share. It's useful for visualising a percentage breakdown or comparing two groups side-by-side.
What is a truncated axis and why is it misleading?
A truncated axis is one that doesn't start at zero. This makes small differences between bars look much larger than they really are. It's a classic technique used in misleading graphs โ always check whether the vertical axis starts at zero.
How do I compare two groups of different sizes?
Convert raw frequencies to percentages within each group first. If 30 out of 60 Year 7s prefer soccer and 10 out of 60 Year 11s do, the percentages are 50 percent versus about 17 percent. Comparing raw numbers when totals differ is misleading.
Video Lesson
- Stats Honors: 1.2 Displaying Categorical Data Watch
Practice Questions
20 questions available.
Practice Questions