SAT Graph and Data Interpretation Questions: A Worked Guide (2026)

SAT graph and data questions test more than reading a point. You may need to compare categories, calculate a rate, interpret slope, connect a table to a written claim, or decide whether a conclusion goes beyond the data. These tasks can appear in Math and in Reading and Writing, but the core habit is the same: identify exactly what the visual measures before doing any calculation.

College Board places ratios, rates, percentages, one- and two-variable data, probability, inference, and statistical claims within Math’s Problem-Solving and Data Analysis domain. Reading and Writing’s Information and Ideas domain can ask you to use a table, bar graph, or line graph as quantitative evidence.

The five-step graph-reading routine

1. Read the title and population

Ask what was measured and for whom. “Average weekly exercise among surveyed students” is not the same as “exercise among all teenagers.”

2. Read both axes and units

Note whether the scale uses raw counts, percentages, thousands, seconds, dollars, or another unit. Check whether the axis begins at zero and whether intervals are even.

3. Identify the requested relationship

Is the question asking for a value, difference, percent change, slope, prediction, association, or conclusion?

4. Calculate only what is needed

Write the mathematical setup before touching the calculator. Many graph errors are task errors, not arithmetic errors.

5. Match the answer’s scope to the evidence

Reject options that change the population, variable, time period, direction, or strength of the claim.

Worked example 1: percent change from a table

Suppose a table shows the number of library visits at two branches:

Branch	April visits	May visits
North	800	920
South	600	720

Question: By what percentage did North Branch visits increase from April to May?

The increase is 920 - 800 = 120. Percent change uses the original value as the denominator:

120 / 800 × 100% = 15%.

The fact that South increased by 120 visits too does not mean both branches had the same percentage increase. South’s base is smaller, so its increase is 120/600=20%.

Write “change ÷ original” before substituting. This prevents the common mistake of dividing by the final value.

Worked example 2: interpreting slope with units

Imagine a line of best fit models plant height h, in centimeters, after d days:

h = 1.8d + 7.5.

The slope is 1.8 centimeters per day. In context, the model predicts that plant height increases by about 1.8 centimeters for each additional day in the observed range. The intercept, 7.5 centimeters, is the predicted height at day 0.

Do not say the slope is “1.8 plants,” and do not automatically treat the association as proof that elapsed time alone caused every change. Keep variables and units attached to numbers.

For a deeper treatment of scatterplots, rates, and inference, use our SAT Problem-Solving and Data Analysis guide.

Worked example 3: quantitative evidence in Reading and Writing

Suppose a passage claims that a new reminder system improved on-time assignment submission in all three classes in a pilot. The table reports:

Class	Before reminders	After reminders
A	72%	81%
B	68%	75%
C	84%	83%

The table does not support the claim that all three classes improved: Class C decreased by one percentage point. A strong answer may say the results were mixed, even though two classes improved.

This question requires claim checking, not averaging. An option stating that the system improved the overall average might be numerically possible, but it would not rescue the exact “all three” claim.

Bar graphs: compare the correct quantities

For bar graphs, first distinguish counts from proportions. If School X has 200 club members and School Y has 120, X has more members. But if the schools have 1,000 and 400 students respectively, the participation rates are 20% and 30%; Y has the higher proportion.

Watch clustered bars. A legend may separate years, groups, or experimental conditions. Trace the correct color or pattern back to the legend before calculating.

If a vertical axis is truncated, the visual difference may look dramatic. Use the labeled values, not the apparent bar height, to compute the actual difference or percent change.

Line graphs: separate value, change, and rate

A point gives a value at a particular input. A vertical difference gives change in output. Slope gives change in output per unit change in input.

Suppose revenue rises from 40,000 in year 1 to 52,000 in year 4. The total change is 12,000 across three yearly intervals, so the average rate of change is `12,000/3 = $4,000 per year`. Dividing by four years would count points rather than intervals.

A line graph that rises throughout an interval shows positive change, but not necessarily a constant rate. A curve becoming steeper has an increasing rate; a straight line has a constant slope.

Scatterplots: association is not causation

A scatterplot can suggest positive association, negative association, or little association. A line of best fit summarizes a trend, and residuals describe differences between observed and predicted values.

Unless the study design supports a causal conclusion, do not claim that one plotted variable causes the other. An observational association between hours of sleep and test performance could reflect other variables. A randomized experiment provides stronger causal evidence, but conclusions still belong to the studied conditions and population.

Be cautious with extrapolation. A model fitted for ages 12–18 may not be reliable at age 60. SAT answers often test whether a prediction remains within the observed data range.

Two-way tables and conditional percentages

In a two-way table, the denominator depends on the condition. If 30 of 120 bus riders prefer an early start, then 25% of bus riders prefer it. If the question asks what percentage of early-start supporters ride the bus, the denominator is all early-start supporters, not all bus riders.

Rewrite “among” or “given that” as the denominator group. Then calculate:

desired intersection / conditioned group.

This habit prevents numerator-denominator reversals in conditional probability.

Statistical claims: ask how the data were produced

Before accepting a conclusion, inspect:

• whether the sample was random or self-selected;
• whether treatments were randomly assigned;
• whether a control or comparison group exists;
• sample size and margin of error;
• whether the conclusion matches the measured variable;
• whether the claim is causal or associational.

If a website asks visitors to click a poll, the results may describe respondents but not a representative population. If participants are randomly assigned to treatments, a difference can support a causal interpretation under the study conditions.

When Desmos helps—and when it does not

The embedded calculator can graph equations, inspect intersections, and work with tables. It can help fit a visible model or verify a line. But it cannot decide which population a claim describes or which denominator the wording requires.

Use Desmos after writing the quantitative target. For a percent problem, enter (new-original)/original, not two numbers without a model. For a line, verify the variables and units before interpreting slope.

Our SAT Math practice-test guide can help you combine calculator work with non-calculator reasoning.

A targeted practice plan

Day 1: complete ten graph/table questions untimed and tag each error as labels, setup, calculation, or scope.

Day 2: drill rates, ratios, and percentages from tables.

Day 3: practice lines, scatterplots, slope, and model interpretation.

Day 4: practice Reading and Writing quantitative-evidence questions.

Day 5: mix all formats under a short time limit.

Day 6: retest the two most common error types with fresh official items.

Use our targeted digital SAT question-bank guide to build filtered sets without repeating the same examples.

Official resources

• College Board’s Problem-Solving and Data Analysis overview lists the current quantitative skills.
• College Board’s Reading and Writing overview explains quantitative command-of-evidence tasks.
• The official Student Question Bank provides filterable Math and Reading and Writing questions.

This independent Makon guide uses original examples. Verify current test details and practice with official College Board materials.