These Digital SAT questions show a scatterplot or its line of best fit and ask you to describe the trend, estimate a value, or decide whether a linear or nonlinear model fits the data.
A scatterplot plots one variable against another so you can see how they move together. A line of best fit is the single straight line that best summarizes that pattern — you can read its slope and y-intercept off the graph (or off two points it clearly passes through) exactly like any other line.
Use Trend → Fit → Estimate: first read the trend (does y increase or decrease as x increases, and how tightly do the points hug that pattern?). Second, find the equation of the line of best fit from two points it passes through. Third, plug in the x-value the question asks about to estimate y.
A scatterplot of test score (y) versus study hours (x) has a line of best fit passing through (2, 66) and (8, 90). Walk through building the equation and using it below.
Using the line of best fit y = 4x + 58 (test score vs. study hours), what score is predicted for a student who studies 3 hours?
With the equation in hand, estimating is just substitution. The next skill is reading a scatterplot's trend and strength without needing an equation at all.
Direction tells you whether the variables move the same way (positive association — points trend upward left to right) or opposite ways (negative association — points trend downward). Strength tells you how closely the points hug that trend: tightly clustered points near a line show a strong association; widely scattered points around the same general trend show a weak one.
These two ideas are independent — a scatterplot can have a clear negative direction but weak strength (points trend down but are scattered), or a clear positive direction with strong strength (points sit almost exactly on an upward line).
Scatterplot A shows points tightly clustered along a downward-sloping line. Scatterplot B shows points loosely scattered but also trending downward. Which shows a stronger negative association?
Once you have the equation, plugging in an x-value gives an estimate for y (or vice versa, solving for x). But not every estimate is equally trustworthy: interpolation — estimating within the range of x-values the data actually covers — is reliable. Extrapolation — estimating outside that range — assumes the trend continues unchanged, which often isn't true.
Before trusting an estimate, check whether the requested x-value falls inside or outside the data's range.
The line of best fit was built from data between x = 10 and x = 50. Classify each estimate.
Now apply the equation itself to a fresh scenario, staying within the data's range.
A scatterplot of daily ice cream sales (y, in dozens) versus high temperature (x, in °F) has a line of best fit y = 1.5x − 60, valid for temperatures between 60°F and 95°F. Predict sales when the temperature is 80°F.
Not every scatterplot is best described by a straight line. If the points curve — rising slowly then sharply, or forming a U-shape — a linear model will systematically miss the pattern. A telltale sign of nonlinear data: the amount y changes per unit of x isn't constant; it grows (or shrinks) as x increases.
College Board lists fitting quadratic and exponential models to scatterplot data, and comparing linear and exponential growth, as part of this skill. Start by recognizing the shape, then match it to the right model in the choices: steadily accelerating growth suggests an exponential model; a dip-then-rise or rise-then-dip shape suggests a quadratic model.
A scatterplot plots a plant's height (cm) against weeks since planting. For the first few weeks the height barely increases, then it grows by more and more each week. Which type of model best fits this pattern?
These mix reading a line of best fit, judging trend and strength, and spotting linear vs. nonlinear shape — the three skills that cover almost every scatterplot question.
Line of best fit: y = 3x + 12. Estimate y when x = 6.
Scatterplot points cluster tightly along an upward line. Describe the association.
A line of best fit was built from data between x = 0 and x = 50. Is estimating at x = 200 interpolation or extrapolation?
A scatterplot's points form a U-shape, decreasing then increasing. What kind of model fits best?
Line of best fit: y = −2x + 90. Estimate y when x = 10.
Points scattered with no visible pattern show what kind of association?
Once these feel automatic, the same center-and-spread instincts you built for distributions transfer directly — you're now comparing how two variables move together instead of the shape of one data set.
Here are five practice problems that help you hone your skills. Use the same method we learned earlier above to solve these problems. Remember, practice makes perfect.
A line of best fit for dollars saved (y) versus weeks (x) is y = 5x + 20. Estimate the savings after 8 weeks.
Scatterplot M's points fall almost exactly on a downward line. Scatterplot N's points trend downward but are loosely scattered. Which shows a stronger negative association?
A line of best fit was built from data between x = 5 and x = 25. Which of the following estimates is an extrapolation?
A scatterplot shows a car's braking distance (y) versus speed (x). As speed increases, braking distance increases by larger and larger amounts. Which model fits best?
A line of best fit for battery percentage (y) versus hours used (x) is y = −1.2x + 100. Estimate the battery percentage after 15 hours.
Related: Distributions, center & spread · Probability · Data inference & margin of error · SAT Math overview