Linear Relationship in AP Statistics

In AP Statistics, a linear relationship is an association between two quantitative variables whose scatter plot follows a roughly straight-line pattern, so a change in x goes with a fairly constant change in y. Its direction and strength are measured by the correlation coefficient, r.

What is Linear Relationship?

A linear relationship means the points on a scatter plot cluster around a straight line. As the explanatory variable (x) goes up, the response variable (y) tends to go up or down by about the same amount each time. That steady rate of change is what makes it linear.

When you describe one, AP Stats wants four things. Give the direction (positive or negative), the form (linear here), the strength (how tightly points hug the line), and any unusual features like outliers. Think of "linear" as the shape label that unlocks everything else. Once you've seen a straight-line pattern, it makes sense to compute r and fit a regression line. If the pattern curves, those tools can mislead you.

Why Linear Relationship matters in AP® Statistics

This term lives in Topic 5.2 Correlation, and it's the foundation the whole topic rests on. The correlation coefficient r measures only linear association. So r is meaningless unless you've first checked that the relationship actually looks like a line. The same check comes up again when you fit and judge a least-squares regression line, because a line is only a good model for linear data. On the exam, "linear" is the form word you need in any description of a scatter plot. It's also the condition you check before you trust r or a regression equation.

How Linear Relationship connects across the course

Correlation Coefficient (Unit 5)

r is basically a score for how linear a relationship is. A value near 1 or -1 means points sit close to a line, and a value near 0 means no linear pattern. A strong curve can still give an r near 0, so r only measures straight-line patterns.

Scatter Plot (Unit 5)

The scatter plot is where you actually see whether a relationship is linear. Always look at the graph before trusting any number. Two data sets can share the same r and look completely different.

Regression Line (Unit 5)

A least-squares regression line is a linear relationship turned into a prediction tool. The slope tells you the predicted change in y for each one-unit increase in x. That only makes sense if the data really follow a line.

Influential Points and Outliers (Unit 5)

A single point can make a relationship look more or less linear than it really is. If removing one point drops r from 0.72 to 0.45, that point was inflating the apparent strength of the linear pattern.

Is Linear Relationship on the AP® Statistics exam?

Multiple choice. You'll interpret r in context and know its limits. For example, an r of -0.85 between study time and anxiety means a strong negative linear association. It does not mean studying causes lower anxiety. Another common stem asks what must be true when r = 0. The answer is that there's no linear relationship, not that there's no relationship at all. You may also see the formula r = (1/n-1)Σ[(xᵢ-x̄)/sₓ][(yᵢ-ȳ)/s_y] and need to interpret the result. Questions about removing an influential point test whether you understand how one value can change the apparent strength of a linear pattern.

Free response. Released FRQs often open with two quantitative variables and ask you to describe or model the relationship. Examples include 2018 Q1 (customers in a checkout line vs. time to finish checkout), 2022 Q1 (bullfrog length vs. mass), and 2026 Q6 (a baseball relationship tied to scoring runs). Full credit usually needs direction, form, strength, and unusual features, all in context with the actual variable names. Writing "linear" without the other parts, or skipping the context, is a common way to lose points.

Linear Relationship vs Nonlinear (curved) relationship

Both are associations between two quantitative variables. The difference is the shape. A linear relationship follows a straight line, and a nonlinear one follows a curve, like a U-shape or a leveling-off pattern. The trap is that r measures only linear strength. A perfect curve can produce an r close to 0, so a small r never proves the variables are unrelated. Always check the scatter plot (and a residual plot, if you've fit a line) before calling something linear.

Key things to remember about Linear Relationship

• A linear relationship means the scatter plot of two quantitative variables follows a roughly straight-line pattern.

• When you describe a scatter plot, include direction, form, strength, and unusual features, and always use the context of the variables.

• The correlation coefficient r measures only the strength and direction of a linear relationship, so r near 0 does not mean there is no relationship.

• A strong linear relationship does not prove causation, even when r is close to 1 or -1.

• A single influential point can make a relationship look much more or much less linear, so check how r changes when it is removed.

• Fitting a regression line only makes sense when the relationship is actually linear.

Frequently asked questions about Linear Relationship

What is a linear relationship in AP Statistics?

It's an association between two quantitative variables whose scatter plot follows a roughly straight line. You measure its direction and strength with the correlation coefficient r, and you model it with a regression line.

Does r = 0 mean there is no relationship between the variables?

No. An r of 0 means there's no linear relationship. The variables could still have a strong curved relationship that r can't detect, which is why you should always look at the scatter plot.

How is a linear relationship different from correlation?

A linear relationship is the straight-line pattern itself. Correlation (r) is the number that measures how strong and in what direction that pattern is. You need the linear pattern first, or r isn't meaningful.

Does a strong linear relationship mean x causes y?

No. Even an r of -0.85 between study time and anxiety only shows a strong negative linear association. Proving causation takes a well-designed experiment, not just a correlation from observational data.

How do I describe a linear relationship on an AP Stats FRQ?

State the direction, form, strength, and any unusual features, and name the actual variables. For bullfrog data like 2022 FRQ Q1, that means talking about length and mass specifically, not just "x and y."