# Comparisons of the Distributions for One Quantitative Variable

> AP Statistics · AP Stats 2024-2026
> Source: https://www.owlsprep.com/study/ap-statistics-u10-comparisons-of-the-distributions-for/

This module teaches you to systematically compare univariate quantitative distributions across groups, following AP Statistics free response scoring rules to maximize points.

**Prerequisites:** [Calculating summary statistics for quantitative data (mean, median, IQR, range)](https://www.owlsprep.com/study/ap-statistics-u10-summary-statistics-quantitative/); [Interpreting basic graphical displays (histograms, boxplots, stemplots)](https://www.owlsprep.com/study/ap-statistics-u10-graphs-for-quantitative-data/)

## Learning objectives

- Identify the 4 mandatory SCSO features to compare across two or more univariate quantitative distributions
- Select AP-approved graphical displays for valid cross-group distribution comparisons
- Write comparison statements that meet College Board free response rubric requirements for full credit
- Detect and avoid misleading comparisons caused by mismatched axes, unequal bins, or unequal sample sizes

## Core SCSO Comparison Framework

Every valid AP distribution comparison must explicitly address four linked features, rather than describing each group's data separately. Vague statements like 'Group A is spread out' will not earn full credit, as they do not reference a second group for contrast.

> **Exam Mnemonic**
>
> SCSO: Shape, Center, Spread, Outliers — tick off all four features before submitting your response to avoid losing easy points.

**Valid Distribution Comparison** — A structured analysis that uses comparative language to contrast all four SCSO features across every group in the dataset, with supporting numerical or graphical evidence.

**Worked example:** Compare the age distributions of 20 physical therapy patients and 25 general wellness patients, given: therapy group median 47, mean 45, IQR 12, no outliers, roughly symmetric; wellness group median 32, mean 38, IQR 18, one high outlier at 71, right skewed.

1. 1. Compare shape: The physical therapy age distribution is roughly symmetric, while the general wellness age distribution is right-skewed.
2. 2. Compare center: The median age for the physical therapy group (47) is 15 years higher than the median age for the wellness group (32).
3. 3. Compare spread: The wellness group ages are more spread out, with an IQR of 18 compared to the therapy group's IQR of 12.
4. 4. Compare outliers: The wellness group has one high outlier at age 71, while the physical therapy group has no identified outliers.

**Check your understanding**

Test your understanding of valid comparison statements

1. Which of the following earns full AP credit?

   - Group A is skewed right. Group B has a median of 22.
   - Group A has a higher median than Group B, and Group B has a larger IQR.
   - Group A's data is more spread out.
   - Group B has an outlier.

   *Why:* This statement explicitly contrasts features across both groups, which is required for full rubric points.

## AP-Approved Graphical Displays for Comparison

Only three graphical displays are accepted by AP graders for comparing univariate quantitative distributions. Using any other display (such as separate disconnected histograms) will result in lost points for the comparison component of the question.

**Comparing methods**

Use this guide to select the correct display for your use case:

- **Side-by-side boxplots** — Multiple boxplots plotted on a single shared numerical axis, ideal for 3+ groups and large datasets
  - Pros: Easy to compare center and spread across many groups, no clutter for large n
  - Cons: Hides fine-grained shape details

- **Back-to-back stemplots** — Shared central stem for two groups, leaves extend left and right, ideal for small datasets (n < 50)
  - Pros: Preserves individual data points, shows exact shape
  - Cons: Impractical for more than 2 groups or large n

- **Overlapping histograms** — Two histograms with distinct color shading on the same axis, ideal for showing fine shape differences
  - Pros: Full visibility of distribution shape
  - Cons: Unreadable with more than 2 groups

**Worked example:** Select the most appropriate graphical display to compare final exam scores across 4 different AP Statistics classes, each with 80 students.

1. 1. Eliminate unsuitable options: Back-to-back stemplots only support 2 groups and small sample sizes, so they are invalid here.
2. 2. Eliminate overlapping histograms: 4 overlapping histograms would be unreadable due to clashing shading and clutter.
3. 3. Confirm side-by-side boxplots: They fit 4 groups easily on a single axis, and allow clear comparison of median, IQR, and outliers across all classes.

**Exam command terms**

The most common command term for this topic has strict AP requirements:

- **Compare the distributions** — You must explicitly contrast all 4 SCSO features across groups, not describe each distribution separately *(Do not write 'Group A is skewed left. Group B is skewed right.' Instead write 'Group A is skewed left while Group B is skewed right.')*

## Avoiding Misleading Comparisons

Even if you follow the SCSO framework, your comparison will be invalid if your underlying graphical or numerical setup is flawed. The most common sources of misleading comparisons are unequal bin widths, mismatched axis scales, and comparing raw counts across groups of very different sizes.

> **warning**
>
> If histograms use different bin widths for each group, any shape or spread comparison you make will be invalid, and you will lose all associated points on the AP exam.

**Worked example:** A student compares two histograms of test scores: Group 1 uses bin widths of 5 points, Group 2 uses bin widths of 10 points. The student concludes Group 2 has a more spread out distribution. Explain why this conclusion is not justified.

1. 1. Identify the flaw: Group 2's bins are twice as wide as Group 1's, so the visual spread of Group 2's histogram is artificially inflated.
2. 2. Correct the setup: Rescale both histograms to use identical 5-point bin widths, and convert counts to relative frequencies if group sample sizes differ.
3. 3. Reassess: After standardization, Group 1's actual IQR is 12, while Group 2's IQR is only 9, so the original conclusion was the opposite of the true result.

**Check your understanding**

Identify the invalid comparison

1. Which of the following comparisons is not justified?

   - Comparing two side-by-side boxplots on the same 0-100 axis
   - Comparing two histograms with different bin widths
   - Comparing two back-to-back stemplots for n=30 each
   - Comparing two distributions using median and IQR

   *Why:* Unequal bin widths create a distorted visual that cannot be used to compare shape or spread fairly.

## Common pitfalls

- **Wrong:** Describing each distribution separately without linking features across groups
  - Why it fails: AP rubrics award zero points for 'parallel descriptions' that do not make explicit cross-group contrasts
  - Correct: Use comparative language like 'higher than', 'more spread out than', 'while' to contrast every SCSO feature across groups
- **Wrong:** Omitting one of the 4 SCSO features from your comparison
  - Why it fails: Most AP free response questions on this topic are scored with 4 holistic points, one for each feature
  - Correct: Use the SCSO mnemonic to tick off shape, center, spread, outliers before you finish writing
- **Wrong:** Using mean to compare center when one distribution has extreme outliers
  - Why it fails: Mean is pulled by outliers, so it is not a representative measure of center for skewed distributions with outliers
  - Correct: Use median to compare center for distributions with outliers or strong skew
- **Wrong:** Comparing raw counts instead of relative frequencies when group sample sizes are very different
  - Why it fails: A larger group will naturally have higher counts in every bin, making the distribution look artificially different
  - Correct: Convert all counts to relative frequencies (proportions) before comparing distributions of unequal size
- **Wrong:** Claiming a small visible difference in distributions is meaningful with no supporting evidence
  - Why it fails: AP graders deduct points for overstating conclusions without supporting numerical or graphical evidence
  - Correct: Qualify claims by referencing the exact difference in median, IQR, or outlier position to back up your statement

## Cheatsheet

| Feature | Required Comparison Action | AP Full Credit Example |
| --- | --- | --- |
| Shape | Explicitly contrast shape across all groups | Group 1 is roughly symmetric, while Group 2 is strongly right-skewed |
| Center | Compare median or mean with explicit values | The median value for Group 1 (38) is 12 units higher than the median for Group 2 (26) |
| Spread | Compare IQR, range, or standard deviation | Group 2 has a larger spread, with an IQR of 17 compared to Group 1's IQR of 9 |
| Outliers | Note presence/absence or location of outliers | Group 1 has two low outliers, while Group 2 has no identified outliers |

## What's next

Mastering distribution comparison is a foundational skill that appears on nearly every AP Statistics exam, most often as the first free response question. The SCSO framework you learned here will help you avoid losing 2-3 easy points that many students miss due to incomplete or non-comparative writing. This descriptive comparison skill directly leads into inferential statistics, where you will learn to quantify whether observed differences in distribution center are statistically significant rather than just descriptive. You will now build on this knowledge to run formal hypothesis tests for two groups, and practice applying your comparison skills to full past AP FRQ sets.

---

From [OwlsPrep](https://www.owlsprep.com) — free study guides for A-Level, IB, AP and IGCSE, written against the official syllabus. Canonical page: https://www.owlsprep.com/study/ap-statistics-u10-comparisons-of-the-distributions-for/