Scatterplots & Association — I can create and interpret scatterplots and describe the direction, form, and strength of association. - Math 1

SECTION 1

Historical Context & Motivation

Humans have always looked for patterns—in the stars, in crop yields, in markets. But for most of history, people relied on intuition rather than data to spot connections between two quantities. The development of the scatterplot gave us a powerful visual tool to examine whether two variables move together, move apart, or have no relationship at all. Understanding these relationships is the foundation of modern data science, medical research, economics, and everyday decision-making.

1686 • First Known Scatter Diagram

English astronomer Edmund Halley plotted barometric pressure against altitude to study atmospheric behavior, creating one of the earliest examples of a scatterplot.

1833 • Statistical Graphics Emerge

John Herschel plotted orbital data for double stars and fit a smooth curve through the points, foreshadowing the idea of trend lines and regression.

1885 • Galton's Regression Studies

Sir Francis Galton used scatterplots to study the relationship between parents' and children's heights, coining the term 'regression' and establishing scatterplots as a core statistical tool.

1896 • Pearson's Correlation Coefficient

Karl Pearson formalized the correlation coefficient (r), giving statisticians a numerical way to measure the strength and direction of a linear relationship seen in scatterplots.

2000s+ • Big Data & Interactive Plots

Software tools like Excel, Desmos, and Python libraries make scatterplots instantly accessible, allowing anyone to explore millions of data points and spot patterns in real time.

The central question that scatterplots help us answer is: When one variable changes, does the other variable tend to change in a predictable way? If so, we say the two variables are associated. Learning to create and read scatterplots is the first step toward answering that question with evidence rather than guesswork.

SECTION 2

Core Principles & Definitions

Before you can interpret a scatterplot, you need to understand a few foundational ideas. A scatterplot displays bivariate data—data that involves two quantitative variables measured on the same individuals or objects. Each data point is plotted as a dot whose horizontal position represents one variable and whose vertical position represents the other. By examining the overall pattern of the dots, you can describe the association between the variables using three key properties: direction, form, and strength.

Direction

A positive association means both variables increase together. A negative association means one increases while the other decreases.

Form

The shape of the overall pattern: linear (points follow a straight-line trend), nonlinear (curved pattern), or no clear pattern.

Strength

How closely the data points follow the trend. Strong associations have tightly clustered points; weak associations have widely scattered points.

Outliers

Individual points that fall far from the overall pattern. Outliers deserve special attention because they can reveal data errors, unusual cases, or important phenomena.

Explanatory vs. Response

The explanatory variable (x-axis) is what you think may influence the response variable (y-axis). Choosing correctly matters for interpretation.

✦

KEY TAKEAWAY

Think of a scatterplot like a bird's-eye view of a crowd at a concert. If everyone is streaming toward the stage (one direction, tightly packed), you see a strong, positive, linear pattern. If people are milling around randomly, there's no association. Describing a scatterplot means reporting the crowd's general direction, whether they move in a straight line or a curve, and how tightly they cluster together.

SECTION 3

Visual Explanation — Reading a Scatterplot

The diagram below shows a scatterplot of study hours versus exam scores for 15 students. Each dot represents one student. The horizontal axis (x-axis) shows the explanatory variable (hours studied), and the vertical axis (y-axis) shows the response variable (exam score). Notice how the dots generally rise from left to right, suggesting that more study time is associated with higher scores.

Each cyan dot represents one student's study hours and exam score. The dashed pink line shows the overall trend. The red dot labeled Outlier falls well below the pattern—perhaps this student was ill on test day.

When you look at this scatterplot, describe it using three words: direction (positive—the points go up from left to right), form (linear—the points follow a roughly straight-line path), and strength (strong—the points are clustered tightly around the trend line). Always mention any outliers you see, because they often tell an interesting story.

SECTION 4

Mathematical Framework — Measuring Association

While describing direction, form, and strength with words is important, statistics also provides a numerical measure of linear association: the correlation coefficient (r). This single number captures both the direction and the strength of a linear relationship. It ranges from −1 to +1, and knowing its formula helps you understand what the number really tells you.

CORRELATION COEFFICIENT

r = (1 / (n − 1)) × Σ [(xᵢ − x̄) / sₓ] × [(yᵢ − ȳ) / sᵧ]

Where n = number of data points, x̄ and ȳ are the means of x and y, and sₓ and sᵧ are the standard deviations of x and y. Each term in the sum compares how far a point is from the mean in standardized units.

In Math 1, you usually won't compute r by hand—your calculator or software does it for you. What matters is understanding what the value tells you. Here are the key interpretive guidelines.

INTERPRETING r

−1 ≤ r ≤ 1

If r > 0, the association is positive (both variables increase together). If r < 0, the association is negative (one increases as the other decreases). Values close to ±1 indicate a strong linear relationship; values near 0 indicate a weak or nonexistent linear relationship.

⚠

Important Caution

The correlation coefficient r only measures linear association. A dataset with a perfect U-shaped curve could have r ≈ 0 even though there is a clear relationship. Also, correlation does not imply causation—just because two variables move together does not mean one causes the other.

SECTION 5

Detailed Breakdown — Types of Association

Not every scatterplot looks the same. The diagram below illustrates four common patterns you will encounter. By recognizing these patterns quickly, you can describe the association and begin to think about what type of model (if any) would fit the data.

Four classic scatterplot patterns. Strong positive linear (top left) and strong negative linear (top right) are well-captured by r. The nonlinear curved pattern (bottom left) has a near-zero r despite a clear pattern—r only measures linear relationships. The no association plot (bottom right) shows random scatter with no discernible pattern.

Pattern	Direction	Form	Strength	Typical r Value
Strong Positive Linear	Positive	Linear	Strong	0.8 to 1.0
Weak Positive Linear	Positive	Linear	Weak	0.3 to 0.5
Strong Negative Linear	Negative	Linear	Strong	−0.8 to −1.0
Nonlinear (Curved)	Varies	Curved	r is misleading	Can be near 0
No Association	None	No pattern	None	Near 0

Summary of common scatterplot patterns and their characteristics

When you describe a scatterplot on a test or assignment, you should always mention all three properties—direction, form, and strength—plus any notable outliers. A complete response might read: "The scatterplot shows a strong, positive, linear association between hours studied and exam score, with one outlier that scored unusually low for its study time."

SECTION 6

Worked Example — From Data to Description

Let's walk through a complete example. A teacher collects data on the number of absences and the final grade (out of 100) for eight students. We'll create a scatterplot, describe the association, and estimate the correlation.

Student	Absences (x)	Final Grade (y)
A	1	95
B	2	90
C	4	82
D	5	78
E	7	70
F	8	65
G	10	55
H	12	50

Absences vs. Final Grade data for 8 students

Creating and Interpreting a Scatterplot

1
Step 1 — Identify Variables

The explanatory variable is the number of absences (x-axis) because we suspect absences influence grades. The response variable is the final grade (y-axis).

x = Absences, y = Final Grade
2
Step 2 — Set Up Axes

Choose scales that cover the range of each variable. Absences range from 1 to 12, so the x-axis should go from 0 to at least 13. Grades range from 50 to 95, so the y-axis should go from about 40 to 100. Label both axes clearly and include units.

x-axis: 0–14 (Absences), y-axis: 40–100 (Final Grade)
3
Step 3 — Plot the Points

For each student, locate the x-value on the horizontal axis and the y-value on the vertical axis, then place a dot at their intersection. Student A with 1 absence and a grade of 95 goes in the upper-left area. Student H with 12 absences and a grade of 50 goes in the lower-right area.

8 data points plotted
4
Step 4 — Describe Direction

As the number of absences increases (moving right), the final grade decreases (moving down). This means the association is negative.

Direction: Negative
5
Step 5 — Describe Form

The points appear to follow a straight-line path rather than curving. This means the form is linear.

Form: Linear
6
Step 6 — Describe Strength & Outliers

The data points are fairly close to where a trend line would be drawn, with no points far from the pattern. The association is strong. There are no apparent outliers. A calculator gives r ≈ −0.99, confirming a very strong negative linear association.

Strong, negative, linear association with no outliers (r ≈ −0.99)

SECTION 7

Strengths & Limitations of Scatterplots

Scatterplots are one of the most useful tools in statistics, but they do have limitations. Understanding both sides helps you know when a scatterplot is the right choice and when you need additional tools.

Strengths	Limitations
Reveal the shape, direction, and strength of a relationship at a glance	Only display two variables at a time (three with color coding)
Make outliers immediately visible	Can become cluttered ('overplotting') with very large datasets
Work for both linear and nonlinear patterns	Require both variables to be quantitative—categorical data needs other graphs
Help you decide whether a linear model is appropriate before fitting a line	Showing an association does NOT prove causation—lurking variables may be involved
Easy to create by hand or with technology	Visual interpretation can be subjective—two people may disagree on 'moderate' vs. 'strong'

Strengths and limitations of scatterplots

✦

KEY TAKEAWAY

A scatterplot is like a weather radar—it gives you a quick, visual overview of what's happening. But just as radar can't tell you why a storm formed, a scatterplot can't tell you why two variables are related. It shows association, not causation. To investigate causes, you need controlled experiments or deeper statistical analysis.

SECTION 8

Connection to Regression & Advanced Topics

In Math 1, you learn to describe association by looking at scatterplots and computing r. In later courses—Algebra 2, AP Statistics, or college-level statistics—you'll build on this foundation in powerful ways. The most important next step is linear regression, which finds the equation of the best-fit line (called the least-squares regression line) through the data. That line lets you make predictions—for example, predicting a student's test score based on how many hours they studied.

What You Learn Now (Math 1)	What Comes Next
Create scatterplots by hand or with technology	Use software to generate regression equations automatically
Describe direction, form, and strength	Compute residuals to assess how well a model fits the data
Interpret r as a measure of linear association	Use r² to quantify the percentage of variation explained by the model
Identify outliers visually	Classify outliers as high leverage, influential points, or both
Recognize linear vs. nonlinear patterns	Apply transformations (log, square root) to linearize curved data

Math 1 foundations vs. advanced statistics concepts

Everything you learn about scatterplots now—choosing the right axes, recognizing patterns, identifying outliers—directly transfers to these advanced topics. Think of this lesson as laying the groundwork for a whole toolkit of statistical analysis methods that you'll encounter in future math and science courses.

SECTION 9

Practice Problems

Question 1: CONCEPTUAL

A scatterplot shows points that fall from left to right and cluster tightly around an imaginary straight line. Describe the association using the three key properties (direction, form, and strength). Would you expect r to be closer to −1, 0, or +1?

Show answer

The association is negative (falling from left to right), linear (follows a straight-line path), and strong (points cluster tightly around the line). You would expect r to be close to −1, because the relationship is both negative and strong. A value like r ≈ −0.90 or lower would be typical for this pattern.

Question 2: BASIC CALCULATION

A dataset has a correlation coefficient of r = 0.72. Describe what this tells you about the direction and strength of the linear relationship. If you added one extreme outlier far from the trend, would r likely increase, decrease, or stay the same?

Show answer

A correlation of r = 0.72 indicates a moderately strong, positive, linear association. Both variables tend to increase together, and the points are reasonably close to a line but not perfectly aligned. Adding an extreme outlier far from the trend would typically decrease the absolute value of r because the outlier pulls the correlation away from the existing linear pattern, increasing the scatter around the trend line.

Question 3: INTERMEDIATE

A researcher plots the age of a car (in years) on the x-axis and the car's resale value (in dollars) on the y-axis. The scatterplot shows a curved pattern: values drop steeply for newer cars but level off for older cars. Should the researcher report the correlation coefficient r as a complete summary of this relationship? Explain your reasoning.

Show answer

No, the researcher should not rely solely on r. The correlation coefficient measures only linear association, and this data has a curved (nonlinear) pattern. Using r alone would understate the strength of the relationship because a straight line is a poor fit for curved data. The researcher should describe the form as nonlinear, note that values decrease quickly at first and then level off, and consider a curved model (such as an exponential decay) rather than a linear one.

Question 4: APPLIED

A health study records the daily step count and resting heart rate for 50 participants. The scatterplot shows a moderate, negative, linear association with r = −0.58 and one outlier—a participant with a very high step count but an unusually high heart rate. Provide a complete description of the scatterplot. Then suggest a real-world explanation for the outlier.

Show answer

The scatterplot shows a moderate, negative, linear association between daily step count and resting heart rate (r = −0.58). As step count increases, resting heart rate tends to decrease. The points are somewhat scattered around the trend, which is why the strength is moderate rather than strong. There is one outlier: a participant with a high step count but an unusually high resting heart rate. A possible explanation is that this person may have an underlying medical condition (such as a thyroid disorder or heart arrhythmia) that keeps their resting heart rate elevated regardless of physical activity, or they may have consumed caffeine or been stressed when the measurement was taken.

Question 5: CRITICAL THINKING

A news article claims: 'Towns with more ice cream shops have higher crime rates, so ice cream shops cause crime.' The scatterplot of number of ice cream shops vs. crime rate shows a strong, positive, linear association with r = 0.85. Explain why the article's conclusion is flawed. Identify at least one lurking variable, and explain how scatterplots should and should not be used in this context.

Show answer

The article commits the classic error of confusing correlation with causation. While the scatterplot correctly shows that more ice cream shops and higher crime rates tend to appear together (r = 0.85), this does not mean one causes the other. A lurking variable is population size: larger towns naturally have both more ice cream shops and more total crimes. Another lurking variable is warm weather, which brings more people outdoors (buying ice cream and creating more opportunities for crime). The scatterplot is correctly used to display and describe the association, but it cannot establish cause and effect. To test causation, you would need a controlled experiment—which is impractical here—or a more sophisticated statistical analysis that accounts for confounding variables.

SUMMARY

Lesson Summary

Summary

A scatterplot displays bivariate data by plotting each data pair as a point on a coordinate plane. The explanatory variable goes on the x-axis and the response variable goes on the y-axis. To describe the association, report three properties: direction (positive or negative), form (linear or nonlinear), and strength (strong, moderate, or weak), along with any outliers.

The correlation coefficient (r) quantifies the direction and strength of a linear relationship on a scale from −1 to +1. Values near ±1 indicate strong linear association, while values near 0 indicate weak or no linear association. Always remember: r only measures linear patterns, and correlation does not imply causation. Scatterplots are the essential first step in understanding the relationship between two quantitative variables, laying the foundation for linear regression and more advanced statistical modeling.