SAT Math Data Representation and Interpretation
Reading charts, tables, and inferring about populations — with no surprise formulas
This topic is mostly careful reading with numbers attached. The SAT tests whether you can read a chart, table, or boxplot and draw a conclusion. Expect 3–4 questions per section, usually in a word context — surveys, studies, population statistics.
The traps are subtle. The SAT routinely confuses correlation with causation, describes random versus non-random samples, and tests whether you can interpret a confidence interval. Read carefully — who studied what, on whom — and you'll catch most of it.
What the SAT actually tests
- Reading values from bar, pie, and scatter charts
- Interpreting box plots (median, quartiles, outliers)
- Computing probabilities from frequency tables
- Inferring about a population from a sample — randomness boundaries
- Confidence intervals and margin of error in context
- Correlation versus causation
Key concepts
Random sample
Conclusions extend to the broader population only when the sample was random. A survey of "students in algebra class" can only describe that class, not all students.
Margin of error
A confidence interval says: "with 95% certainty, the true value lies in ." Larger sample → smaller margin.
Correlation vs causation
Two variables can move together without one causing the other. The classic SAT example: ice cream sales and drownings. Both rise in summer, but ice cream doesn't cause drownings.
Frequency table
Counts observations per category. Probability = count for the category / total count.
Worked examples
A survey of 200 randomly selected city residents found that 60% support a park project. Margin of error: . What does this conclude?
The true population support lies (with high confidence) in , i.e., between 56% and 64%. The sample value is an estimate, not exact.
💡 Margin of error creates an interval around the estimate. Don't treat the sample number as exact.
In a frequency table of 100 students, 30 walk, 50 take the bus, 20 are driven. What is the probability that a randomly chosen student takes the bus?
.
💡 Probability from a table = category count / total.
Common pitfalls
- Generalizing beyond the studied population — from "high schoolers" to "all Poles."
- Confusing correlation with causation. "X and Y rise together" doesn't mean "X causes Y."
- Ignoring margin of error — treating the sample value as exact.
- Reading from a chart without checking units (thousands? percent?). Easy mistake to make under time pressure.
Exam strategy
Read the question twice. First — what does the chart or table show? Second — what specifically is asked? Note units (thousands, percents) and which group was studied. For "what can we conclude?" questions, apply two filters: (1) was the sample random? (2) is the population in question the same one the sample came from? If not, the answer is usually "we cannot conclude."
Frequently asked questions
When can I generalize a sample result to a population?
Only when the sample was random AND the target population matches the one the sample was drawn from. Otherwise the answer is "we cannot conclude."
How do I read a boxplot?
Box = IQR (Q1 to Q3). Middle line = median. Whiskers = min/max (or up to 1.5×IQR). Points beyond whiskers = outliers.
Observational study vs experiment?
An experiment randomly assigns participants to groups (control vs treatment). An observational study just observes existing differences. Only experiments support causal conclusions.
What is a confidence interval?
A range that contains the true parameter value with a stated probability (usually 95%). Narrower = more precise. Larger sample size usually gives a narrower interval.
Practice 60+ data-interpretation questions