Segmented Bar Charts — I can represent categorical data with segmented bar charts and interpret comparisons. - Math 1

SECTION 1

Historical Context & Motivation

Humans have been trying to make sense of data for centuries, and one of the most persistent challenges has been figuring out how to show the composition of something — not just its total size, but what it's made of. Before modern charts existed, people relied on tables full of numbers, which could be overwhelming and hard to compare at a glance. The development of bar charts and their segmented variations gave us a visual shortcut for understanding how parts relate to wholes.

The story of segmented bar charts is intertwined with the broader history of statistical graphics. As societies grew more complex — tracking trade, populations, and public health — the need for clear, comparative visuals became urgent. Pioneers in data visualization realized that stacking colored segments inside a single bar could reveal patterns that raw numbers never could.

1786 • William Playfair Invents the Bar Chart

Scottish engineer William Playfair published the first known bar chart in his Commercial and Political Atlas, using bars to compare Scotland's trade data. This laid the groundwork for all bar-based visualizations.

1858 • Florence Nightingale's Rose Diagrams

Nightingale used segmented polar-area diagrams to show causes of soldier deaths during the Crimean War. Her work demonstrated that stacking categories within a single graphic could persuade policymakers to act.

1920s • Stacked Bars Enter Textbooks

By the early twentieth century, stacked (segmented) bar charts became a standard tool in statistics textbooks, used to display census data, economic indicators, and survey results.

2000s–Present • Digital Tools Make Segmented Charts Universal

Software like Excel, Google Sheets, and programming libraries such as R and Python's matplotlib have made creating segmented bar charts accessible to students, journalists, and researchers worldwide.

The central question that segmented bar charts answer is deceptively simple: How do the proportions of different categories compare across groups? Whether you're comparing survey responses from two schools, transportation preferences across cities, or the composition of diets across countries, segmented bar charts turn complex categorical data into an intuitive visual story.

SECTION 2

Core Principles & Definitions

Before constructing a segmented bar chart, you need to understand a few foundational ideas. A segmented bar chart (also called a stacked bar chart) divides each bar into colored segments that represent subcategories. The length of each segment is proportional to the count or percentage of that subcategory. When every bar is scaled to the same total height (100%), the chart becomes a 100% segmented bar chart, which is specifically designed for comparing proportions rather than raw counts.

Categorical Data

Data that can be sorted into groups or categories — such as favorite color, mode of transportation, or grade level — rather than measured on a numerical scale.

Two-Way Table

A table that displays the frequency (count) of data for two categorical variables simultaneously. It is the data source from which a segmented bar chart is built.

Relative Frequency

The proportion of a specific subcategory within a group, calculated by dividing the subcategory count by the group total. This is what determines segment length in a 100% segmented bar.

Conditional Distribution

The distribution of one categorical variable within a fixed level of another variable. A 100% segmented bar chart visually displays the conditional distribution for each group.

Association

Two categorical variables are associated if the conditional distribution of one variable changes across levels of the other. If the segments look roughly the same across bars, there is little to no association.

✦

KEY TAKEAWAY

Think of a segmented bar chart like a stack of colored building blocks inside each bar. If you're comparing two groups — say, freshmen and seniors — each bar is a tower built with the same set of colored blocks. A 100% segmented bar chart makes every tower the same height so you can focus entirely on whether the color ratios differ. If the stacks look identical, the two groups have similar distributions. If the color bands shift dramatically, there's an association between the variables.

SECTION 3

Visual Explanation

The diagram below shows a 100% segmented bar chart comparing how students in three different schools get to school. Each bar represents one school, and the colored segments within each bar represent the proportion of students who walk, bike, take the bus, or drive. Because every bar reaches 100%, we can directly compare proportions across schools.

Each bar reaches 100%, making it easy to compare the conditional distribution of transportation mode across schools. Notice how School B has a much larger "Walk" segment (cyan) and a smaller "Bike" segment (violet) compared to the others, suggesting different commuting patterns.

Looking at the chart, you can immediately spot differences. School B has the highest proportion of walkers at 40%, while School C relies most heavily on the bus at 40%. School A's segments are the most evenly distributed. These visual differences in the segment widths are exactly what allow us to identify associations between two categorical variables — in this case, the school a student attends and how they commute.

SECTION 4

Mathematical Framework

Building a 100% segmented bar chart requires converting raw frequency counts into relative frequencies (proportions or percentages). The formulas are straightforward, but understanding what each calculation represents is essential for interpreting the chart correctly.

RELATIVE FREQUENCY (PROPORTION)

Relative Frequency = (Subcategory Count) ÷ (Group Total)

This gives a decimal between 0 and 1. Multiply by 100 to convert to a percentage. For example, if 24 out of 80 students in School A walk, the relative frequency is 24 ÷ 80 = 0.30, or 30%.

SEGMENT HEIGHT (IN A 100% BAR)

Segment Height = (Relative Frequency) × (Total Bar Height)

If the total bar height on your chart is 10 cm and the relative frequency is 0.30, then the segment's height is 0.30 × 10 = 3 cm. Every segment's height is determined this way, and they must sum to the total bar height.

VERIFICATION CHECK

∑ (All Relative Frequencies within a group) = 1.00 (or 100%)

Always verify that your proportions add up to 1 (or 100%). If they don't, you've made an arithmetic error or are missing a category.

💡

Counts vs. Percentages

A segmented bar chart using raw counts will have bars of different heights (proportional to group size). A 100% segmented bar chart standardizes every bar to the same height. The 100% version is preferred for comparing proportions, especially when group sizes are unequal. If School A surveyed 80 students and School B surveyed 200, raw-count bars would make School B's bar much taller, obscuring the proportional differences.

SECTION 5

Interpreting Segmented Bar Charts

Creating a segmented bar chart is only half the task — the real skill lies in interpreting it. When you look at a 100% segmented bar chart, you should focus on three things: individual segment sizes, differences between corresponding segments across bars, and the overall pattern of whether the bars look similar or different.

When the colored bands are nearly the same size across all bars (left), there is no association between the variables. When the bands shift dramatically (right), an association is present. Always ask the three questions shown at the bottom.

When interpreting a segmented bar chart in a statistics class, you should use specific, comparative language. Instead of saying "School B has a lot of walkers," say "The proportion of students who walk is highest at School B (40%), compared with 30% at School A and 20% at School C." Providing percentages and naming the groups strengthens your interpretation and shows statistical reasoning.

⚠

Beware of Misleading Comparisons

If you use a segmented bar chart with raw counts instead of percentages and the groups are very different sizes, the visual can be misleading. A bar for a group of 500 people will tower over a bar for a group of 50, even if their proportional breakdowns are identical. Always consider whether a 100% segmented bar chart is more appropriate for the comparison you want to make.

SECTION 6

Worked Example

A school surveyed 200 students about their favorite type of music. The results are broken down by grade level in the two-way table below. Let's walk through how to construct and interpret a 100% segmented bar chart from this data.

Music Genre	9th Grade (n = 60)	10th Grade (n = 50)	11th Grade (n = 45)	12th Grade (n = 45)
Pop	24	15	9	5
Hip-Hop	18	15	13	14
Rock	12	10	14	18
Country	6	10	9	8

Two-way table: Music preference by grade level

Constructing and Interpreting a 100% Segmented Bar Chart

1
Step 1 — Compute Relative Frequencies for Each Grade

Divide each count by the grade's total. For 9th grade: Pop = 24 ÷ 60 = 0.40 (40%), Hip-Hop = 18 ÷ 60 = 0.30 (30%), Rock = 12 ÷ 60 = 0.20 (20%), Country = 6 ÷ 60 = 0.10 (10%). Check: 40% + 30% + 20% + 10% = 100%. ✓

9th grade: Pop 40%, Hip-Hop 30%, Rock 20%, Country 10%
2
Step 2 — Repeat for Remaining Grades

10th grade: Pop = 15 ÷ 50 = 30%, Hip-Hop = 15 ÷ 50 = 30%, Rock = 10 ÷ 50 = 20%, Country = 10 ÷ 50 = 20%. 11th grade: Pop = 9 ÷ 45 = 20%, Hip-Hop = 13 ÷ 45 ≈ 28.9%, Rock = 14 ÷ 45 ≈ 31.1%, Country = 9 ÷ 45 = 20%. 12th grade: Pop = 5 ÷ 45 ≈ 11.1%, Hip-Hop = 14 ÷ 45 ≈ 31.1%, Rock = 18 ÷ 45 = 40%, Country = 8 ÷ 45 ≈ 17.8%.

All four grades' relative frequencies verified to sum to 100%.
3
Step 3 — Draw the Bars

Create four bars of equal height, one for each grade. Stack the genre segments within each bar using the computed percentages. Assign a consistent color to each genre across all bars (for example, Pop = cyan, Hip-Hop = violet, Rock = amber, Country = pink). Label the axes: the horizontal axis shows grade level, and the vertical axis shows percentage from 0% to 100%.
4
Step 4 — Interpret the Chart

Look for how the segments shift across the bars. The Pop segment shrinks as grade level increases (from 40% in 9th grade to about 11% in 12th grade). The Rock segment grows (from 20% in 9th grade to 40% in 12th grade). Hip-Hop remains fairly stable around 29–31%. Country varies without a clear trend.

There appears to be an association between grade level and music preference. Pop music's share decreases with grade level, while Rock music's share increases.

SECTION 7

Strengths, Limitations & Chart Comparisons

Segmented bar charts are one of several tools for displaying categorical data. Understanding when to use them — and when another chart type would be better — is an important part of statistical literacy. The table below compares four common chart types for categorical data.

Chart Type	Best For	Limitations
100% Segmented Bar Chart	Comparing proportions of subcategories across groups when group sizes differ.	Hides actual counts; middle segments are harder to compare because they don't share a common baseline.
Side-by-Side Bar Chart	Comparing individual category counts or proportions with a common baseline for each segment.	Uses more horizontal space; harder to see the overall composition of each group.
Pie Chart	Showing proportions of a single group's composition.	Difficult to compare multiple groups; humans are poor at judging angles.
Mosaic Plot	Showing both the group sizes and proportions simultaneously in a single chart.	More complex to read; not commonly covered in introductory courses.

Comparison of chart types for categorical data

✦

WHEN TO CHOOSE A SEGMENTED BAR CHART

Choose a 100% segmented bar chart when your primary goal is comparing the proportional makeup of several groups. Think of it like comparing the recipes of two different smoothies: you don't care that one person made 32 ounces and the other made 16 — you care whether the banana-to-strawberry ratio is different. If you also need to compare the actual counts, consider a regular segmented bar chart or a side-by-side bar chart instead.

SECTION 8

Connection to Advanced Statistical Concepts

Segmented bar charts are your first window into a bigger statistical idea: testing for independence between two categorical variables. In later courses, you'll learn formal methods — like the chi-square test of independence — that quantify whether the differences you see in a segmented bar chart are statistically significant or could have occurred by chance alone.

Concept	What You're Learning Now	Where It Leads
Conditional distribution	Reading and comparing the proportions within each bar of a segmented bar chart.	Calculating expected counts under independence for chi-square tests.
Association	Visually noticing when segment proportions differ across groups.	Using a chi-square statistic to measure the strength of association numerically.
Two-way tables	Organizing data and computing row/column percentages.	Computing expected frequencies and test statistics for hypothesis testing.
Relative frequency	Converting counts to proportions for chart construction.	Connecting to probability: relative frequency as an estimate of theoretical probability.

How segmented bar chart skills connect to advanced statistics

For now, the key insight is that your eyes are doing informally what the chi-square test does formally. When you look at a 100% segmented bar chart and notice that the segments are dramatically different across bars, you're making an informal judgment that the variables are associated. Future courses will give you the mathematical tools to back up that judgment with a p-value and a confidence level.

SECTION 9

Practice Problems

Question 1: CONCEPTUAL

A 100% segmented bar chart shows data from two cities. All four colored segments appear to be roughly the same width in both bars. What does this tell you about the relationship between city and the categorical variable represented by the colors?

Show answer

When the segments are roughly the same width in both bars, the conditional distributions are approximately equal across the two cities. This means there is little to no association between the city and the categorical variable. In other words, knowing which city a person is from does not help you predict their category — the proportional breakdown is essentially the same in both places.

Question 2: BASIC CALCULATION

A survey of 120 students at Riverside High asks about their favorite lunch option. The results are: Pizza = 48, Salad = 30, Sandwich = 24, Other = 18. Calculate the relative frequency (percentage) of each category and verify they sum to 100%.

Show answer

Divide each count by the total of 120. Pizza: 48 ÷ 120 = 0.40 = 40%. Salad: 30 ÷ 120 = 0.25 = 25%. Sandwich: 24 ÷ 120 = 0.20 = 20%. Other: 18 ÷ 120 = 0.15 = 15%. Verification: 40% + 25% + 20% + 15% = 100%. ✓ These percentages would determine the height of each segment in a 100% segmented bar chart for Riverside High.

Question 3: INTERMEDIATE

Two stores track customer payment methods. Store A (n = 200): Cash 60, Credit 80, Debit 40, Mobile Pay 20. Store B (n = 150): Cash 15, Credit 45, Debit 45, Mobile Pay 45. Compute the relative frequencies for each store and describe what a 100% segmented bar chart would reveal about the association between store and payment method.

Show answer

Store A: Cash = 60 ÷ 200 = 30%, Credit = 80 ÷ 200 = 40%, Debit = 40 ÷ 200 = 20%, Mobile Pay = 20 ÷ 200 = 10%. Store B: Cash = 15 ÷ 150 = 10%, Credit = 45 ÷ 150 = 30%, Debit = 45 ÷ 150 = 30%, Mobile Pay = 45 ÷ 150 = 30%. The bars would look quite different. Store A's bar is dominated by Credit (40%) and Cash (30%), while Store B's bar is more evenly split among Credit, Debit, and Mobile Pay (30% each) with a much smaller Cash segment (10%). This indicates an association between the store and payment method preference.

Question 4: APPLIED

A health researcher surveys teens in urban and rural areas about their primary form of exercise. Urban (n = 250): Team Sports 75, Gym 100, Running 50, None 25. Rural (n = 200): Team Sports 60, Gym 20, Running 80, None 40. Construct the conditional distributions, describe what the 100% segmented bar chart would look like, and discuss whether exercise type appears to be associated with location.

Show answer

Urban: Team Sports = 75 ÷ 250 = 30%, Gym = 100 ÷ 250 = 40%, Running = 50 ÷ 250 = 20%, None = 25 ÷ 250 = 10%. Rural: Team Sports = 60 ÷ 200 = 30%, Gym = 20 ÷ 200 = 10%, Running = 80 ÷ 200 = 40%, None = 40 ÷ 200 = 20%. The chart would show two bars of equal height. Team Sports would be roughly the same size in both bars (30%), but the Gym segment would be dramatically larger in the Urban bar (40% vs. 10%) and the Running segment would be much larger in the Rural bar (40% vs. 20%). The None segment would also differ (10% vs. 20%). Because the conditional distributions shift substantially between the two groups, there appears to be an association between exercise type and location.

Question 5: CRITICAL THINKING

A classmate creates a segmented bar chart (not a 100% chart) using raw counts to compare pet ownership across three neighborhoods. Neighborhood A has 500 residents, Neighborhood B has 80 residents, and Neighborhood C has 300 residents. The classmate concludes that Neighborhood A has the strongest preference for dogs because its dog segment is the tallest. Explain why this conclusion may be flawed and propose a better approach.

Show answer

The conclusion is flawed because the group sizes are vastly different. Neighborhood A has 500 residents — more than six times the size of Neighborhood B — so its bar will naturally be much taller overall, making every segment (including dogs) appear larger in absolute terms. The tall dog segment may simply reflect the large population, not a stronger preference. A better approach is to create a 100% segmented bar chart, which would normalize each bar to the same height and allow a fair comparison of proportions. For example, if 200 out of 500 residents in Neighborhood A own dogs (40%) and 40 out of 80 in Neighborhood B own dogs (50%), Neighborhood B actually has a higher proportion of dog owners despite having a shorter raw-count segment. This illustrates why proportional comparisons require standardized charts.

SUMMARY

Lesson Summary

Summary

A segmented bar chart divides each bar into colored segments representing subcategories of categorical data. When standardized to equal heights, the chart becomes a 100% segmented bar chart that shows the conditional distribution of one variable within each level of another. Building one requires computing relative frequencies from a two-way table and verifying that proportions within each group sum to 100%.

To interpret a segmented bar chart, compare the segment widths across bars. If the segments look roughly the same, there is no association between the variables. If the segments shift noticeably, an association exists. This visual skill lays the groundwork for formal tests of independence, like the chi-square test, that you will encounter in more advanced statistics courses. Always use a 100% chart when comparing groups of different sizes, and describe patterns with specific percentages and group names.