Review AP Statistics Unit 2 to build your skills in analyzing relationships between two variables, from two-way tables and scatterplots to correlation, least squares regression, residuals, and data transformations. This unit bridges descriptive statistics and the inference work that runs through Units 6 through 9.
Use the topic guides, key terms, and practice questions available for this unit to work through every concept from 2.1 to 2.9.
Unit 2 asks you to move beyond single-variable summaries and investigate whether two variables are related. You will work with both categorical and quantitative data, using different tools for each type.
Two-way tables organize counts or relative frequencies for two categorical variables. You calculate joint, marginal, and conditional relative frequencies, then use side-by-side bar graphs, segmented bar graphs, or mosaic plots to compare distributions and judge whether an association exists.
Scatterplots display bivariate quantitative data. You describe them using form, direction, strength, and unusual features, then quantify the linear association with the correlation coefficient r, which is unit-free and always between -1 and 1.
The least squares regression line (LSRL) uses ŷ = a + bx to predict responses. You interpret slope and intercept in context, use r² to measure how much variation is explained, check residual plots for linearity, and apply transformations when the relationship is curved.
Even a strong correlation coefficient close to 1 or -1 does not mean one variable causes the other. Lurking variables, confounding, and random variation can all produce apparent associations. This principle appears throughout the AP Statistics exam and is essential for interpreting any two-variable analysis.
Frames the unit by asking whether apparent patterns between two variables are real or due to random variation. Introduces explanatory and response variables and the distinction between observational studies and experiments.
Covers two-way tables, joint relative frequencies, and graphical displays including side-by-side bar graphs, segmented bar graphs, and mosaic plots for comparing two categorical variables.
Focuses on calculating and comparing marginal and conditional relative frequencies from a two-way table to determine whether two categorical variables are associated.
Introduces scatterplots for bivariate quantitative data and the four-part description framework: form, direction, strength, and unusual features.
Defines the correlation coefficient r, its formula, its properties (unit-free, range -1 to 1), and the critical principle that correlation does not imply causation.
Introduces the regression equation ŷ = a + bx, how to calculate predicted values, how to interpret slope and intercept in context, and the risks of extrapolation.
Defines residuals as y - ŷ and explains how residual plots are used to assess whether a linear model is appropriate for a data set.
Covers the LSRL formulas for slope (b = r(sy/sx)) and intercept (a = y-bar - b(x-bar)), the meaning of r², and how to interpret regression coefficients in context.
Identifies outliers, high-leverage points, and influential points in regression, and introduces log and power transformations to linearize curved data sets.
Open this guide for a closer review of the topic.
Open this guide for a closer review of the topic.
Open this guide for a closer review of the topic.
Open this guide for a closer review of the topic.
Open this guide for a closer review of the topic.
Open this guide for a closer review of the topic.
Open this guide for a closer review of the topic.
Open this guide for a closer review of the topic.
Open this guide for a closer review of the topic.
Open this guide for a closer review of the topic.
Open this guide for a closer review of the topic.
This snapshot uses Fiveable practice activity to show where students tend to miss questions and which review moves are worth prioritizing first.
Across 8.1k multiple-choice practice attempts for this unit.
Practice activity included in this snapshot.
Across 23 scored free-response attempts for this unit.
Review Statistics for Two Categorical Variables with attention to how the concept appears in AP-style source and evidence questions.
Review Representing Two Categorical Variables with attention to how the concept appears in AP-style source and evidence questions.
Review Least Squares Regression with attention to how the concept appears in AP-style source and evidence questions.
Review Representing the Relationship Between Two Quantitative Variables with attention to how the concept appears in AP-style source and evidence questions.
Topic 2.1 frames the entire unit: when you observe an apparent pattern between two variables, you must ask whether it reflects a real association or just random variation. The type of variables determines which tools you use next.
Two-way tables (contingency tables) display counts or relative frequencies for two categorical variables. The three types of relative frequency serve different purposes: joint frequencies describe individual cells, marginal frequencies describe one variable overall, and conditional frequencies let you compare distributions across groups to detect association.
| Frequency type | Denominator | What it tells you |
|---|---|---|
| Joint relative frequency | Grand total | Proportion of all observations in that cell |
| Marginal relative frequency | Grand total | Overall distribution of one variable |
| Conditional relative frequency | Row or column total | Distribution of one variable within a subgroup |
A scatterplot places the explanatory variable on the x-axis and the response variable on the y-axis, with one dot per observation. Every scatterplot description must address four features: form, direction, strength, and unusual features such as outliers or clusters.
The correlation coefficient r quantifies the direction and strength of the linear association between two quantitative variables. It is calculated as r = (1/(n-1)) x sum of [(xi - x-bar)/sx times (yi - y-bar)/sy], but technology is used in practice. Key properties: r is unit-free, always between -1 and 1, and sensitive to outliers.
The regression equation ŷ = a + bx predicts the response variable from the explanatory variable. The slope b is the predicted change in y for each one-unit increase in x; the intercept a is the predicted y when x = 0, which sometimes has no meaningful interpretation. A residual is the difference between the actual value and the predicted value: residual = y - ŷ. Residual plots are used to check whether a linear model is appropriate.
| Residual pattern | Interpretation |
|---|---|
| Random scatter around zero | Linear model is appropriate |
| Curved (U-shape or arch) | Nonlinear model may fit better |
| Fan shape (spread increases) | Variability is not constant across x-values |
The LSRL minimizes the sum of squared residuals and always passes through the point (x-bar, y-bar). The slope formula b = r(sy/sx) connects correlation directly to the regression line. The coefficient of determination r² tells you the proportion of variation in y that is explained by the linear relationship with x.
Not all data sets are well described by a linear model. Topic 2.9 covers two situations: unusual points that distort the LSRL, and curved relationships that require transformation before fitting a line.
Try AP-style multiple-choice questions and written prompts after you review the notes.
| Term | Definition |
|---|---|
| Explanatory Variable | The variable used to explain or predict the response variable; placed on the x-axis of a scatterplot. |
| Response Variable | The outcome variable being predicted or measured; placed on the y-axis of a scatterplot. |
| Two-Way Table | A table that displays counts or relative frequencies for two categorical variables simultaneously, also called a contingency table. |
| Conditional Relative Frequency | A cell frequency divided by its row or column total; used to compare distributions of one categorical variable within subgroups of another. |
| Marginal Relative Frequencies | Row or column totals in a two-way table divided by the grand total; describe the overall distribution of one variable regardless of the other. |
| Scatterplot | A graph that displays two quantitative variables as ordered pairs, with the explanatory variable on the x-axis and the response variable on the y-axis. |
| Correlation Coefficient | The value r that measures the direction and strength of the linear association between two quantitative variables; unit-free and always between -1 and 1. |
| Correlation Does Not Imply Causation | A strong r value does not establish that changes in x cause changes in y; lurking variables or confounding may explain the association. |
| Simple linear regression | A model of the form ŷ = a + bx that uses one explanatory variable to predict a response variable using the least squares criterion. |
| Residual Plot | A graph of residuals (y - ŷ) plotted against x or ŷ; a random scatter with no pattern supports the use of a linear model. |
| Coefficient of Determination | r², the proportion of variation in the response variable explained by the linear relationship with the explanatory variable. |
| Influential Point | A data point whose removal substantially changes the slope, intercept, or correlation of the regression line; often an outlier or high-leverage point. |
| Mosaic Plot | A graphical display for two categorical variables where rectangle areas are proportional to frequencies, making it easy to compare conditional distributions. |
| direction of association | Whether the relationship between two variables is positive (both increase together) or negative (one increases as the other decreases). |
| Bivariate Data | Data collected on two variables for each individual in a sample, used to investigate whether the variables are related. |
A high r value, even r = 0.99, does not mean x causes y. Always note that lurking variables or confounding could explain the association, especially when data come from an observational study.
Joint, marginal, and conditional relative frequencies answer different questions. To check for association between two categorical variables, you need conditional relative frequencies, not marginal ones.
The intercept a is the predicted y when x = 0. If x = 0 is outside the range of the data or has no real-world meaning (for example, a person with 0 hours of sleep), the intercept should not be interpreted literally.
A high r² does not guarantee a linear model is appropriate. Always check the residual plot. A curved pattern in the residuals means the linear model is not the right choice, even if r² looks high.
Plugging an x-value outside the observed range into ŷ = a + bx produces an unreliable prediction. The further you extrapolate, the less trustworthy the result.
AP Statistics consistently asks you to interpret statistical values in the context of the problem, not just state a formula. For slope, intercept, r, and r², your answer must name the actual variables and units. A response that says only 'the slope is 2.3' without context will not receive full credit.
A common task in Unit 2 is deciding whether a linear model is appropriate. You are expected to cite specific evidence: the scatterplot shows a linear form, the residual plot shows no pattern, and r² is reasonably high. For transformed data, you compare residual plots and r² before and after transformation.
Questions about two-variable data often include a follow-up asking whether the association implies causation. The correct response identifies the study design (observational vs. experimental), notes that confounding or lurking variables may be present, and states that causation requires random assignment.
Open the individual guides for Unit 2 when you want a closer review of one topic.
browse guidesPractice free-response reasoning and compare your answer with scoring guidance.
practice FRQsWatch past review streams filtered to Unit 2 when you want a video walkthrough.
open videosUse unit cheatsheets for a quick visual review after you work through the notes.
open cheatsheetsEstimate your broader AP score goal after you review the course and exam format.
open calculatorAP Stats Unit 2 covers 9 topics focused on exploring relationships between two variables using correlation, regression, and data displays. Topics include two-way tables (2.2, 2.3), scatterplots (2.4), correlation (2.5), linear regression models (2.6), residuals (2.7), least squares regression (2.8), and analyzing departures from linearity (2.9). The unit builds from asking whether variables are related all the way to evaluating how well a linear model fits data. See AP Stats Unit 2 for matched practice on each topic.
AP Stats Unit 2 makes up 5-7% of the AP exam. That weight covers everything from correlation and scatterplots to least squares regression and residuals. It's a smaller unit by exam weight, but the skills here, especially interpreting regression output and residual plots, show up repeatedly in later units and in FRQs.
The AP Stats Unit 2 progress check in AP Classroom includes both MCQ and FRQ parts drawn from the unit's 9 topics. MCQ questions test your ability to read scatterplots, interpret correlation, and identify patterns in two-way tables. FRQ questions typically ask you to describe a linear regression model, interpret residuals, or explain why correlation does not imply causation. The progress check pulls heavily from topics 2.4 through 2.9, so make sure you're comfortable with least squares regression and analyzing departures from linearity before you sit for it. Practice with matched questions at AP Stats Unit 2.
AP Stats Unit 2 FRQs most often ask you to interpret a linear regression model, describe what residuals tell you about a model's fit, or explain the meaning of correlation in context. To practice, focus on writing full sentence interpretations of slope, y-intercept, and r-squared values, since partial credit on the AP exam depends on precise wording. Good practice steps: - Sketch and describe residual plots for patterns - Interpret least squares regression equations in context - Explain why a high correlation does not prove causation Find FRQ-style practice questions at AP Stats Unit 2.
You can find AP Stats Unit 2 MCQ and practice test questions at AP Stats Unit 2. That page has multiple-choice questions and practice problems covering all 9 topics, from two-way tables and correlation to residuals and least squares regression. For the best prep, mix MCQ practice with short written responses so you're ready for both question types on exam day.
Start AP Stats Unit 2 by building a solid understanding of correlation before moving into regression. Correlation measures the strength and direction of a linear relationship, but it does not tell you one variable causes another, and that distinction comes up constantly on the exam. A practical study plan: 1. Review two-way tables and practice calculating relative frequencies (topics 2.2-2.3) 2. Sketch scatterplots and describe form, direction, and strength (topic 2.4) 3. Learn what the correlation coefficient r tells you and what it does not (topic 2.5) 4. Practice writing full interpretations of slope and y-intercept for linear regression models (topic 2.6) 5. Read and describe residual plots to check model fit (topics 2.7-2.8) 6. Study patterns like outliers and influential points that signal departures from linearity (topic 2.9) Spend extra time on residuals and regression interpretation since those are the highest-payoff skills for both the progress check and the AP exam. Use AP Stats Unit 2 to find practice matched to each topic.