Skip to content

Unit 5 of 5

Regression Analysis

15% of the exam120 reviewed questions32 key terms16 min read

Regression Analysis is worth about 15% of the AP Statistics exam. Our bank tests it with 120 reviewed questions across 32 distinct ideas — 43 easier, 56 medium and 21 harder.

On this page · 12 sections

This unit is about two quantitative variables measured on the same individuals, such as hours of sleep and reaction time for each student in a class. You start with a picture, the scatterplot, and describe what you see. Then you put a number on the linear part of the pattern with the correlation r. Finally, you fit a line, the least-squares regression line (LSRL), and use it to make predictions.

Every later step depends on an earlier one. The slope of the LSRL is built from r and the two standard deviations, so a mistake about r spreads into the slope. Residuals measure how far each point sits from the line. The residual plot then tells you whether a line was the right model in the first place. R squared tells you how much of the variation in y the line accounts for. None of it means anything unless the form of the scatterplot is roughly linear.

The exam rarely asks you to compute these numbers from raw data, because your calculator or the computer output does that. It asks you to interpret them in context, to pick the right number for a claim, and to say what the numbers cannot tell you. The three big limits are that correlation is not causation, that predictions outside the data are risky, and that a high r squared does not prove a line fits well.

The unit in a minute

One line per topic. Tap any line for the full notes.

  1. Reading a scatterplotDescribe direction, form, strength and unusual features, in context, before you compute anything.
  2. Correlation r and its propertiesr is a unitless measure of linear strength and direction. Shifts, rescaling by a positive constant and swapping the axes do not change it, and multiplying one variable by a negative constant flips its sign.
  3. Building the least-squares regression lineb = r(sy/sx), and the line passes through (x-bar, y-bar); find the slope first, then the intercept.
  4. Interpreting slope and intercept in contextSlope = predicted change in y per one-unit increase in x; intercept = predicted y at x = 0, which may have no real meaning.
  5. ResidualsResidual = observed - predicted; positive means the point sits above the line, so the model underestimated it.
  6. Residual plots and model fitRandom scatter means a line fits; a curve means a line is not appropriate; a fan means predictions vary in precision.
  7. R squared, the coefficient of determinationR squared = share of variation in y explained by the line; it carries no sign and says nothing about form.
  8. Prediction, extrapolation and causationPredict only inside the data range, and never claim cause from observational data, however strong r is.
Grind Regression Analysis questionsAlready know this? Skip the notes and prove it.

The big ideas

Look before you calculate

Correlation and the LSRL only describe linear patterns. A strong curved relationship can produce an r near 0, and a curved pattern can still produce a high r squared. So the scatterplot and the residual plot come first. If the residual plot shows a curve, the linear model is the wrong tool, and its slope, predictions and r squared all lose their meaning.

Each statistic answers one question

r tells you the direction and strength of a linear association. The slope tells you how much predicted y changes when x goes up by one unit. The intercept is the predicted y when x is 0. R squared is the share of the variation in y explained by the line. A residual is one point's prediction error. Most wrong answers in this unit swap one of these for another, such as reading r as a slope or r squared as a correlation.

Everything is said in context

A bare number earns little credit. You must name the variables, give their units, and say "predicted" when you talk about values from the line. Saying that each additional degree lowers the predicted heating bill by $2.50 is a full answer, while saying the slope is -2.5 is not. The model describes averages and predictions, never a guaranteed outcome for one individual.

Association is not causation

A regression line on observational data shows that two variables move together. It cannot show that one causes the other, however strong r is. A lurking variable may drive both, or the link may be coincidence. Only a well-designed randomized experiment supports a cause-and-effect claim. This idea shows up in multiple-choice distractors and in almost every free-response conclusion.

Study notes

8 topics. Each one opens with its key idea, then the detail you need to use it.

1. Reading a scatterplot

Key idea: Describe direction, form, strength and unusual features, in context, before you compute anything.

Put the explanatory variable on the horizontal axis and the response variable on the vertical axis. The explanatory variable is the one you use to predict, and the response is the one you predict. Then describe four things, always in context.

Direction is positive if y tends to rise as x rises, and negative if y tends to fall. Form is linear if the points cluster around a straight line, or nonlinear if they bend, level off or curve back. Strength is how tightly the points hug that form. Unusual features include outliers, gaps and separate clusters.

Form matters most, because everything else in the unit assumes a linear form. Picture car speed against fuel economy. Mileage rises up to about 50 miles per hour and then falls. That relationship is strong, but it is not linear, so a straight line through it would miss badly at both ends.

A complete description names the variables in every part. For engine size and highway mileage, you would report a strong, negative, linear association, plus one car that gets much better mileage than its engine size suggests. Naming the variables turns a generic sentence into an answer that earns credit.

2. Correlation r and its properties

Key idea: r is a unitless measure of linear strength and direction. Shifts, rescaling by a positive constant and swapping the axes do not change it, and multiplying one variable by a negative constant flips its sign.

The correlation r measures the direction and strength of a linear association. It runs from -1 to 1. Values near 1 or -1 mean the points lie close to a line, and values near 0 mean little or no linear association. So r = 0.85 between study hours and test scores describes a strong, positive linear association, while r = -0.12 means there is almost no linear pattern at all.

The exam tests these properties directly. First, r has no units, so changing feet to meters or Fahrenheit to Celsius leaves it alone. Second, adding a constant to every value, or multiplying by a positive constant, does not change r. Multiplying by a negative number flips its sign. Third, r is the same if you swap x and y. Fourth, r only measures linear association, so a perfect U-shape can give r = 0. Fifth, r is not resistant, which means a single outlier can pull it up or down a lot.

The number r is not a slope and not a percentage. If r = 0.85, you cannot say y goes up 0.85 per unit of x, and you cannot say 85% of anything is explained. Both of those claims need other quantities, which the next topics cover.

3. Building the least-squares regression line

Key idea: b = r(sy/sx), and the line passes through (x-bar, y-bar); find the slope first, then the intercept.

The LSRL is written y-hat = a + bx, where y-hat is the predicted value of y. It is the one line that makes the sum of the squared residuals as small as possible. That is why it is called "least squares."

You can build it from five summary numbers. The slope is b = r(sy/sx), with the spread of y on top. The line always passes through the point (x-bar, y-bar), so the intercept is a = y-bar - b(x-bar).

Try it with ages and heights of children. Suppose x-bar = 8 years, sx = 2 years, y-bar = 130 cm, sy = 12 cm and r = 0.9. The slope is 0.9(12/2) = 5.4 cm per year. The intercept is 130 - 5.4(8) = 86.8 cm. So the line is predicted height = 86.8 + 5.4(age).

The common slips are predictable. Using sx/sy gives the slope for predicting x from y, which is a different line. Using r alone as the slope ignores the units of both variables. Forgetting the factor r gives a slope of sy/sx. That slope is at least as steep as the correct one, and it is always positive, so a negative association gets a line that slopes the wrong way.

The formula also explains a hard property question. If you add 10 to every y-value, sy does not change, so the slope stays the same. Only y-bar moves, and that shifts the intercept up by 10.

4. Interpreting slope and intercept in context

Key idea: Slope = predicted change in y per one-unit increase in x; intercept = predicted y at x = 0, which may have no real meaning.

The slope is the predicted change in the response for each one-unit increase in the explanatory variable. A full answer names one more unit of x, says the predicted y rises or falls, and gives the amount b in the units of y, on average. Suppose a line predicts monthly heating cost from outdoor temperature as cost = 150 - 2.5(temperature). The slope means that for each 1-degree rise in temperature, the predicted heating cost drops by $2.50.

Two errors cost points here. One is reversing the variables, for example by claiming that each extra dollar of cost lowers the temperature by 2.5 degrees. The slope always describes y per unit of x, not the other way round. The other is leaving out "predicted" or "on average," which turns a model statement into a promise about every month.

The intercept is the predicted y when x = 0. In the heating line, the predicted cost at 0 degrees is $150. The intercept is not the mean of y. The line passes through y-bar at x-bar, not at x = 0.

Always ask whether x = 0 makes sense. A line predicting weight from height has an intercept for a person 0 inches tall. That is far outside the data, so the intercept is just a number that places the line, not a real prediction.

5. Residuals

Key idea: Residual = observed - predicted; positive means the point sits above the line, so the model underestimated it.

A residual is the error the line makes for one data point. It equals the observed y minus the predicted y, or residual = y - y-hat. To find one, plug the point's x into the line, then subtract that prediction from the actual y.

Here is an example. A line predicts quiz score from minutes of study as score-hat = 52 + 0.6(minutes). A student studied 40 minutes and scored 70. The prediction is 52 + 0.6(40) = 76, so the residual is 70 - 76 = -6 points. The negative sign means the point lies below the line, and the line overestimated her score by 6 points.

The sign carries the meaning. A positive residual means the actual value was higher than predicted, so the point is above the line. A negative residual means the actual value was lower than predicted, so the point is below it. Read the context carefully. For a race time, a positive residual means the runner was slower than predicted, because a bigger time is worse.

Try one where lower is better. A line predicts marathon time from training as time-hat = 300 - 1.2(hours), in minutes. A runner trained 60 hours, so her predicted time is 300 - 72 = 228 minutes. She finished in 236 minutes, so her residual is 236 - 228 = +8 minutes. She ran 8 minutes slower than the model predicted, even though her residual is positive.

For any LSRL, the residuals add up to 0. Points with large residuals are outliers in the y-direction, because the line predicts them poorly. If you get the order backwards and compute predicted minus observed, every sign flips and your interpretation becomes wrong.

6. Residual plots and model fit

Key idea: Random scatter means a line fits; a curve means a line is not appropriate; a fan means predictions vary in precision.

A residual plot puts x (or the predicted values) on the horizontal axis and the residuals on the vertical axis. A horizontal line at 0 marks where a perfect prediction would sit. The plot removes the linear trend so you can see what is left over.

If a line is a good model, the residuals form random scatter with no pattern around 0. A curve, such as a U-shape, means the relationship is nonlinear. The line overpredicts in some regions and underpredicts in others in a systematic way, so a linear model is not appropriate. Note that this is a pattern across all the points, not one strange point. A nonlinear model would describe the data better than a straight line.

A fan shape, where the residuals spread out as x grows, is called heteroscedasticity, or non-constant variance. The line may still follow the trend, but predictions are less precise where the residuals spread out. A single isolated residual far from 0 is an outlier, which is a different problem from a curve.

Put together, a linear model is appropriate when the scatterplot looks roughly linear and the residual plot shows random scatter with roughly constant spread.

Do not confuse a residual plot with a normal probability plot, which only shows whether the residuals look roughly normal and says nothing about whether a line is the right shape.

7. R squared, the coefficient of determination

Key idea: R squared = share of variation in y explained by the line; it carries no sign and says nothing about form.

R squared is the fraction of the variation in the response that the linear model with x explains. If r squared = 0.81 for house price on square footage, then 81% of the variation in house prices is accounted for by the linear relationship with square footage. The other 19% comes from things the model leaves out, such as location or age of the house.

The word "variation" is not optional. R squared does not mean 81% of houses are predicted correctly, and it does not mean prices are 81% accurate. It compares how spread out the residuals are with how spread out y is on its own.

To move between r and r squared, square or take the square root. If r = 0.6, then r squared = 0.36. If r squared = 0.81, then r is either 0.9 or -0.9. You pick the sign from the slope or the scatterplot, because squaring wipes out the direction. A frequent trap reports r = 0.85 and offers "85% of the variation is explained" as an answer. The correct figure is 0.85 squared, which is about 72%.

A high r squared does not prove that a line is the right shape. Curved data can still give r squared = 0.9, so check the residual plot as well. Computer output often also lists s, the typical size of a residual in the units of y.

8. Prediction, extrapolation and causation

Key idea: Predict only inside the data range, and never claim cause from observational data, however strong r is.

To predict, plug an x-value into the line. Use the prediction with care, because the line is only a summary of the data you have.

Extrapolation means predicting for an x-value outside the range of the data. It is unreliable because nothing guarantees the linear pattern continues. A line fitted to children aged 2 to 12 might predict a 40-year-old to be about 3 meters tall. Growth levels off, so the straight line fails. Predictions inside the data range are called interpolation, and they are much safer.

The second limit is causation. A strong correlation in observational data does not show that x causes y. Shoe size and reading score are strongly correlated among elementary students. Bigger feet do not improve reading. Age is a lurking variable that drives both, because older children have bigger feet and read better. When a correlation comes from a lurking variable or from pure coincidence, it is called a spurious correlation.

Only a randomized experiment, where the researcher assigns x, can support a causal claim. A correct conclusion for observational data uses the word "associated," not "causes" or "leads to."

The model's other limitations come from what it leaves out. A line predicting house price from square footage ignores location, so two same-size houses in different towns get the same prediction. More houses make the slope more precise, but they cannot add a variable the model never measured. For the same reason, a bigger observational sample cannot remove a lurking variable. And the line gives the average price for a given size, not the price of one house.

Common mistakes

Drawn from the wrong answers students actually pick on these questions.

Reading r as the slope, as in "r = 0.85, so scores rise 0.85 points per hour," or flipping the ratio in the slope formula to sx/sy.

The slope is b = r(sy/sx), with the spread of the response on top. It carries the units of y per unit of x, while r has no units and only measures linear strength and direction. Using sx/sy gives the line that predicts x from y, which is a different line.

Treating r as the percent of variation explained, or treating r squared as the correlation.

Square r to get the share of variation explained, so r = 0.85 gives about 72%, not 85%. Going the other way, r squared = 0.81 means r = 0.9 or -0.9. Take the sign from the slope.

Mixing up slope and intercept, or calling the intercept the average of y.

The slope is the predicted change in y per one-unit increase in x. The intercept is the predicted y when x = 0. The line passes through (x-bar, y-bar), so the mean of y sits at x-bar, not at x = 0.

Saying r squared = 0.81 means 81% of predictions are accurate.

R squared describes variation, not individual hits. It says 81% of the variation in y is explained by the linear relationship with x. Individual predictions can still miss by a lot.

Using the wrong graph for the question, for example reading a curved residual plot as "outliers" or a normal probability plot as a test of linearity.

A curved pattern across all residuals means a nonlinear relationship, so a line is the wrong model. A fan means non-constant spread. A lone far-off residual is an outlier. To judge whether a line fits, use the residual plot, not a normal probability plot.

How this unit shows up on the exam

Regression is one of the five units on the revised AP Statistics exam, and most of the questions on it test interpretation rather than calculation. Multiple-choice stems usually give you a fitted line, a value of r or r squared, or a piece of computer output, and then ask "Which of the following is the correct interpretation?" The three wrong choices usually come from the mistakes in this guide, such as treating r as a slope, treating r squared as a correlation, reversing x and y, or claiming causation. They are not always that same set, so check every choice against the definitions.

Expect quick calculations too. You may need to build the LSRL from means, standard deviations and r, compute one residual, or turn r into r squared. Property questions ask what happens to r, the slope or the intercept when every value is shifted, rescaled or converted to new units. Picture questions show a residual plot and ask whether a line is appropriate.

Computer regression output is a favorite. Learn to find the slope (the coefficient next to the x-variable's name), the intercept (labeled "Constant" or "Intercept", depending on the software), R-sq, and s. Then write the equation using the real variable names.

On free-response questions, readers reward context and precision. An interpretation of slope earns credit only if it names both variables, gives units, and uses "predicted" or "on average." A judgment about whether a linear model fits must point to the residual plot, not just to r. Questions that ask about causation reward a clear statement that observational data show association only, with a plausible lurking variable named.

How the exam actually asks it

Real items from the AP Statistics bank, easier first. The answer is marked, and every wrong choice is explained — because knowing why a distractor is tempting is most of the work.

Worked example 1Easier
A statistics student calculates a correlation coefficient of r = 0.85 between study hours and exam scores for a sample of 30 students. Which of the following is the most accurate interpretation of this value?
  1. AFor each additional hour of study, exam scores increase by 0.85 points.
  2. B85% of the variation in exam scores is explained by the linear relationship with study hours.
  3. CThere is a strong, positive linear association between study hours and exam scores.Correct
  4. DStudying more hours causes exam scores to increase by a factor of 0.85.

Why C is right

The correlation coefficient r = 0.85 indicates the strength and direction of the linear association. A value of 0.85 is close to 1, indicating a strong positive linear relationship. It does not describe causation, it is not a slope, and r^2 (not r) gives the percent of variation explained.
Why the other 3 choices miss
A
Incorrect because r is not the slope. The slope of the LSRL is b1 = r*(Sy/Sx), which requires the standard deviations of both variables.
B
Incorrect because 85% of variation explained corresponds to r^2, not r. Since r = 0.85, r^2 = 0.7225, meaning about 72.25% of variation is explained.
D
Incorrect because correlation does not imply causation. Additionally, r does not quantify a multiplicative factor.
Worked example 2Easier
A regression model predicts monthly electricity bills (in dollars) from average daily temperature (in degrees Fahrenheit). The LSRL is: Bill = 120 - 0.8*(Temperature). Which of the following is the correct interpretation of the slope?
  1. AThe average electricity bill is $120 more than what temperature alone would predict.
  2. BFor each additional dollar in the electricity bill, the average temperature decreases by 0.8 degrees.
  3. CFor each one-degree increase in temperature, the predicted electricity bill decreases by $0.80, on average.Correct
  4. DWhen the temperature is 0 degrees, the predicted electricity bill is $0.80.

Why C is right

The slope b1 = -0.8 is interpreted as: for each one-unit increase in the explanatory variable (temperature in degrees Fahrenheit), the predicted value of the response variable (bill in dollars) changes by -0.8, meaning it decreases by $0.80 on average.
Why the other 3 choices miss
A
Incorrect because that is a garbled interpretation mixing the intercept concept with incorrect phrasing. The $120 is simply the predicted bill when temperature equals zero.
B
Incorrect because this reverses the roles of the variables. The slope describes the change in the response (bill) per unit change in the explanatory variable (temperature), not the other way around.
D
Incorrect because that is the interpretation of the intercept, not the slope. The intercept (120) represents the predicted bill when temperature is 0.
Worked example 3Easier
A researcher fits a least-squares regression line to predict crop yield (bushels/acre) from amount of fertilizer (pounds/acre). The y-intercept of the LSRL is 45. Which of the following is the best interpretation of this value?
  1. AWhen no fertilizer is applied, the predicted crop yield is 45 bushels per acre.Correct
  2. B45% of the variation in crop yield is explained by the amount of fertilizer applied.
  3. CThe average crop yield in the study was 45 bushels per acre.
  4. DFor each additional pound of fertilizer, the predicted yield increases by 45 bushels per acre.

Why A is right

The y-intercept is interpreted as the predicted value of the response variable when the explanatory variable equals zero. Here, when fertilizer = 0, the model predicts a yield of 45 bushels per acre. Whether this is meaningful depends on whether x = 0 is in the range of the data.
Why the other 3 choices miss
B
Incorrect because the percent of variation explained is r^2, the coefficient of determination, not the intercept value.
C
Incorrect because the y-intercept does not represent the average of y. The regression line passes through (x-bar, y-bar), but the intercept is the predicted y when x = 0.
D
Incorrect because that description defines the slope, not the intercept. The slope tells us how much the response changes per unit increase in the explanatory variable.
Worked example 4Easier
A linear regression model for predicting house price (in thousands of dollars) from square footage produces r^2 = 0.81. Which of the following is the correct interpretation of this value?
  1. AFor each additional square foot of living space, the predicted house price increases by 0.81 thousand dollars.
  2. BThe correlation between house price and square footage is 0.81.
  3. C81% of house prices can be accurately predicted using this regression model.
  4. D81% of the variation in house prices is accounted for by the linear relationship with square footage.Correct

Why D is right

The coefficient of determination r^2 = 0.81 means that 81% of the variability in the response variable (house price) is explained by the linear regression on the explanatory variable (square footage). The remaining 19% is unexplained variability (residual).
Why the other 3 choices miss
A
Incorrect because r^2 is not the slope. The slope requires the ratio of standard deviations multiplied by r.
B
Incorrect because r^2 = 0.81 means r = 0.9 (taking the positive square root since the relationship is presumably positive). The correlation is r = 0.9, not 0.81.
C
Incorrect because r^2 does not tell us the percentage of individual predictions that are accurate. It measures the proportion of total variability in y explained by the model.

Can you do all of these?

Tick what you can do without looking. Anything left unticked is where to go back to.

0 of 10 ticked

Ready to test yourself?

Open practice pulls from the 36 free questions in this unit. The other 84 come with the one-time AP purchase, which unlocks every AP course.

Official topic list

The official course framework breaks this unit into 5 numbered topics.

  1. 5.1Graphical Representations Between Two Quantitative Variables
  2. 5.2Correlation
  3. 5.3Linear Regression Models
  4. 5.4Residuals
  5. 5.5Least-Squares Regression

Key terms in Regression Analysis

Ordered by how often our reviewed questions test them. Open any term to see the questions that use it.

The rest of AP Statistics

Every unit with its share of the exam. You are on Unit 5.

  1. 1Exploring One-Variable Data and Collecting Data25% of exam
  2. 2Probability, Random Variables, and Probability Distributions20% of exam
  3. 3Inference for Categorical Data: Proportions20% of exam
  4. 4Inference for Quantitative Data: Means15% of exam
  5. 5Regression Analysis15% of exam