Statistics for Two Categorical Variables

students, have you ever wondered whether two groups are connected in a meaningful way? 🤔 In this lesson, you will explore how statisticians study two categorical variables at the same time. This matters in real life when we want to compare groups like grade level and sports participation, gender and favorite subject, or whether someone wears a helmet and whether they get injured. The goal is not just to list data, but to see whether an association exists and what the data seem to suggest.

By the end of this lesson, you should be able to:

Define important terms for comparing two categorical variables.
Organize data in a two-way table.
Find and interpret conditional distributions.
Compare groups using proportions and percentages.
Recognize association and understand what it does and does not mean.
Connect this topic to the larger AP Statistics unit on exploring two-variable data.

What Are Two Categorical Variables?

A categorical variable places individuals into groups or categories rather than giving numerical measurements. Examples include $\text{yes/no}$ answers, blood type, eye color, and type of school club. When we study two categorical variables, we are asking whether the distribution of one variable changes across the categories of the other.

For example, suppose a school surveys students about:

whether they play a sport: $\text{Yes}$ or $\text{No}$
whether they are in an honors class: $\text{Yes}$ or $\text{No}$

Now we have two categorical variables, each with two categories. This is a common setting in AP Statistics because it helps us compare groups clearly and draw conclusions from data. 📊

A very important idea is that we are usually looking for association, not necessarily a cause-and-effect relationship. If two variables are associated, the distribution of one variable differs across the categories of the other variable. Association means there is a relationship in the data, but it does not prove that one variable causes the other.

Two-Way Tables: Organizing the Data

The main tool for two categorical variables is the two-way table. A two-way table shows counts for all combinations of the categories.

Here is an example:

|------------|-------------|---------------------|-------|

| Honors | $18$ | $12$ | $30$ |

| Not Honors | $22$ | $28$ | $50$ |

| Total | $40$ | $40$ | $80$ |

This table tells us how many students fall into each combination of categories. For instance, $18$ students are both in honors and play a sport.

Two-way tables help us answer questions like:

What percent of honors students play a sport?
What percent of non-honors students do not play a sport?
Are these percentages similar or different?

When working with a two-way table, always identify the explanatory variable and the response variable if the context suggests one. The explanatory variable is the group or factor used for comparison, and the response variable is the outcome being compared. In AP Statistics, this is especially useful when describing conditional distributions.

Marginal and Conditional Distributions

A marginal distribution describes the overall distribution of one categorical variable without considering the other variable. In the table above, the marginal distribution of honors status is based on the row totals:

Honors: $30$ out of $80$
Not Honors: $50$ out of $80$

So the marginal percentages are:

Honors: $\frac{30}{80} = 0.375 = 37.5\%$
Not Honors: $\frac{50}{80} = 0.625 = 62.5\%$

A conditional distribution describes the distribution of one variable within a particular category of the other variable. For example, what is the distribution of sports participation given that a student is in honors?

For honors students:

Plays sport: $\frac{18}{30} = 0.60 = 60\%$
Does not play sport: $\frac{12}{30} = 0.40 = 40\%$

For non-honors students:

Plays sport: $\frac{22}{50} = 0.44 = 44\%$
Does not play sport: $\frac{28}{50} = 0.56 = 56\%$

This comparison is the heart of the lesson. students, ask yourself: do the percentages look the same across groups? If they do not, that suggests an association between the two variables.

Comparing Groups with Relative Frequencies

In AP Statistics, it is usually better to compare relative frequencies or percentages than raw counts. Why? Because groups may have different sizes. A larger group may naturally have bigger counts even if the pattern is the same.

For the example table, the count of sport players is larger in the non-honors group ($22$ vs. $18$), but the conditional proportion of sport players is actually higher among honors students:

Honors: $60\%$
Not Honors: $44\%$

That is why a good statistical comparison should use percentages, not just counts. 📈

A complete comparison should mention:

Which groups are being compared
The specific percentages or proportions in each group
The direction of the difference
Whether the difference seems important or large

For example: “A greater percentage of honors students play sports ($60\%$) than non-honors students ($44\%$), so the data suggest an association between honors status and sports participation.”

Association, Independence, and Interpretation

Two categorical variables are independent if the conditional distribution of one variable is the same for every category of the other variable. In other words, knowing one variable gives no useful information about the other.

Using the table above, if sports participation were independent of honors status, the percent of students who play sports would be about the same in both groups. But $60\%$ and $44\%$ are not the same, so the variables do not appear independent.

If two variables are not independent, we say they show association. In AP Statistics, you may describe this as:

“There appears to be an association between $X$ and $Y$.”
“The distribution of $Y$ changes across categories of $X$.”

Be careful with wording. Association does not mean one variable causes the other. For example, being in honors may be linked with sports participation, but the table alone does not prove a direct cause. Other variables could be involved, such as time management, family support, or school scheduling.

Graphs for Two Categorical Variables

A bar graph or segmented bar graph is often used to display two categorical variables. These graphs make it easier to compare conditional distributions visually.

A segmented bar graph shows each category of the explanatory variable as a bar with the same total length of $100\%$. Each bar is divided into segments based on the response variable’s proportions. This is especially helpful because it makes differences in group distributions easy to see.

For the sports and honors example, a segmented bar graph would show:

One bar for honors students with $60\%$ playing sports and $40\%$ not playing sports
One bar for non-honors students with $44\%$ playing sports and $56\%$ not playing sports

When two bars look very different, that suggests association. When the bars look nearly the same, that suggests little or no association.

Real-World Meaning and AP Statistics Reasoning

students, one of the most important skills in AP Statistics is connecting numbers to context. A two-way table is not just arithmetic; it is a way to study how groups differ in the real world.

Here are some examples:

A health survey might compare $\text{vaccinated}$ and $\text{hospitalized}$.
A school study might compare $\text{tutoring}$ and $\text{passing a class}$.
A transportation survey might compare $\text{uses public transit}$ and $\text{lives in the city}$.

In each case, the same reasoning applies:

Organize the data in a two-way table.
Compute conditional distributions.
Compare percentages.
Describe any association.
Avoid claiming causation unless the study design supports it.

If the data come from a random sample, then the results may be generalized to the population. If the data come from a randomized experiment, then a cause-and-effect conclusion may be possible. If the data are from an observational study, association can be described, but causation should not be claimed.

Conclusion

Statistics for two categorical variables helps us answer questions about relationships between groups. By using two-way tables, conditional distributions, and graphical displays, students, you can compare categories clearly and identify association when it appears. This topic is a key part of exploring two-variable data because it teaches you how to make sense of patterns when both variables are categorical. It also builds strong AP Statistics habits: use percentages, compare groups carefully, and make conclusions that match the type of data collected. ✅

Study Notes

A categorical variable puts individuals into groups, not numerical measurements.
A two-way table displays counts for all combinations of two categorical variables.
A marginal distribution summarizes one variable overall.
A conditional distribution summarizes one variable within a category of the other variable.
Use relative frequencies or percentages to compare groups, not just counts.
If conditional distributions differ, the variables show association.
If conditional distributions are the same, the variables are independent.
Association does not mean causation.
Segmented bar graphs are useful for visual comparison of categorical data.
In AP Statistics, always connect numerical results to the real-world context.