satified

Data Analysis · Skill 4 of 7

SAT Scatterplots and Two Variable Data Practice

Scatterplot questions on the digital SAT are reading questions wearing math clothes: what does the slope of the fit line mean, what does the line predict, and how far off is the real data point. Master those three reads and this becomes one of the most reliable point sources on the test.

  • Domain: Data Analysis
  • About 15% of the test is Data Analysis
  • Difficulty: easy to hard
  • Free, no account
Drill scatterplots now → See all Data Analysis skills

The patterns the SAT actually uses

Four patterns cover the whole skill. Each one turns on the difference between the data cloud and the line drawn through it.

Pattern 01

Point versus line

Some questions want the actual data point, others want the value the line of best fit predicts at the same x. Deciding which one is being asked is most of the work.

Pattern 02

Interpret slope and intercept

The fit line's slope is the predicted change in y for one more unit of x, in the story's units. The intercept is the predicted value when x is zero, if that even makes sense in context.

Pattern 03

Measure the miss

Actual minus predicted: how far a point sits above or below the line, and whether the line predicts too high or too low there. This is residual thinking without the vocabulary word.

Pattern 04

Name the association

Positive or negative, strong or weak, linear or curved. A rising cloud means positive. A tight curve means strong and nonlinear, not no association, and that distinction gets tested.

One worked example, start to finish

Worked example · medium

A researcher fits the line y = 3.2x + 14 to a scatterplot, where x is the number of weeks a store has been open and y is its average daily customer count. One store has been open for 5 weeks and averages 34 daily customers. By how much does the actual customer count exceed the count predicted by the line?

  1. The predicted value comes from the line: substitute x = 5 into y = 3.2x + 14.
  2. Compute it: 3.2 × 5 = 16, and 16 + 14 = 30, so the line predicts 30 daily customers.
  3. The actual value is the data point itself: 34 daily customers.
  4. Subtract, actual minus predicted: 34 − 30 = 4. The point sits 4 above the line, which is why the question says exceeds.

Answer: 4 customers

The context rotates through plants, sales, and temperatures on regeneration. Predicted from the line, actual from the point, subtract. That is the entire pattern.

Where students lose the point

  • Reporting predicted when asked for actual. The point and the line both live at x = 5. One question in every set hinges on which of the two y values you report, or on their difference.
  • Building slope from data points. Data points are not on the line of best fit. Slope questions about the line must use the line itself, its equation or two points that are actually on it.
  • Extending the line too far. A fit built on weeks 5 to 30 says little about week 200. Choices that extrapolate confidently past the plotted data are bait, and careful language wins.
  • Upgrading association to cause. A tight scatter shows two variables moving together. Nothing in a scatterplot alone shows that one causes the other, and the modest answer choice is the credited one.

Using Desmos here

This is one of the best Desmos skills on the whole test. Put the data points into a table and Desmos plots the scatter. Ask it for a linear regression and the line of best fit appears with its slope and intercept, so predicted values come from one substitution. Satified ships the same built in calculator, so you can rehearse that exact workflow and also learn when a question is pure interpretation and the calculator never needs to open.

Why drilling here is different

Satified's scatterplot drills regenerate the data cloud, the fit line, and the question focus every time they load, from easy reads to hard residual reasoning. That variety trains the three reads this page covers instead of one chart's answers. All 1,483 questions in the bank have independently verified answers and explanations.

A new scatter, every load.

Start this skill free →

Questions students ask

What is a residual on the SAT?
The gap between a real data point and the value the line of best fit predicts at the same x. The SAT rarely uses the word residual; it asks how much the actual value exceeds or falls short of the predicted one.
Do I have to compute the line of best fit?
No. The line is always drawn on the plot or its equation is given. Your jobs are reading predictions off it, interpreting its slope and intercept in context, and comparing it against actual data points.
What does a strong association look like?
Points hugging a clear line or curve. Strength is about tightness and direction is about sign: rising left to right is positive, falling is negative. A tight curve is a strong nonlinear association, which is a favorite trick.
Can Desmos draw a line of best fit?
Yes. Enter the points in a table and ask for a linear regression, and Desmos produces the fitted line instantly. Satified includes the same calculator, so you can practice deciding when the table trick beats eyeballing.
How many scatterplot questions will I see?
Usually 2 to 3 across your two 35 minute modules, inside the Data Analysis domain that covers about 15 percent of SAT Math. They lean heavily toward interpretation rather than computation.

Keep going

Two variable thinking leans on the one variable vocabulary and feeds straight into the probability, inference, and claims skills ahead.