Problem Solving & Data Analysis · ~15% of Math section

SAT Statistical Inference and Study Design: Practice Questions & Study Guide

Questions testing whether you can identify what a study's design allows you to validly conclude, including generalizability and causality.

10 practice questions
3 Easy
4 Medium
3 Hard

Understanding Statistical Inference and Study Design

Statistical inference questions focus on three distinctions: whether a result generalizes beyond the sample, whether a design supports causation or only association, and whether the conclusion matches the study design.

Generalizability depends on sampling. A random sample from a defined population can support a conclusion about that population. A convenience or self-selected sample does not justify extending the result to a broader group without additional evidence.

Causal inference depends on random assignment to conditions. An observational study can establish an association but cannot by itself show that one variable caused another, regardless of sample size. Check answer choices for causal verbs that the design does not support.

A larger random sample generally reduces sampling variability and margin of error, but representativeness still depends on how that sample was selected. Compare both sample size and selection method when evaluating an estimate.

Key Rules & Formulas

Memorize these rules — they come up directly in practice questions.

1

Random sample → can generalize to the population; convenience/volunteer sample → cannot generalize

A randomly selected national sample of 1,000 teens supports conclusions about all teens; a survey of 1,000 volunteers at a mall does not

2

Randomized experiment → can infer causation; observational study → association only

If researchers randomly assign participants to exercise or no-exercise groups and measure health outcomes, they can conclude exercise causes the measured improvement

3

Larger, representative sample → smaller margin of error → more precise estimate

A poll of 2,000 randomly selected voters is more precise than a poll of 200

4

Confounding variable: a third variable that is related to both the independent and dependent variables and could explain the association

People who carry lighters are more likely to get lung cancer—but lighters don't cause cancer; smoking is the confounding variable

5

The scope of a conclusion must match the scope of the sample (group, time period, geography)

A study of 9th-graders at one urban school cannot support conclusions about 'all U.S. high school students'

Statistical Inference and Study Design Practice Questions

Select an answer and click Check Answer to reveal the full explanation. Questions go from easiest to hardest.

Question 1Easy

A researcher surveys 50 randomly selected students at a single high school about their screen time habits. Which of the following conclusions is best supported by the data?

Show explanation

Correct answer: A. The results can reasonably be generalized to students at this school

Explanation

A random sample from a specific school can support conclusions about students at that school. It cannot be generalized nationally without a national sample.

Question 2Easy

A study found that cities with more hospitals have higher rates of disease. A journalist writes, 'Hospitals cause disease.' Which type of error does this conclusion represent?

Show explanation

Correct answer: B. The study confuses correlation with causation

Explanation

An observational study showing an association between two variables does not establish that one causes the other. The more likely explanation is a confounding variable (sicker populations attract more hospitals).

Question 3Easy

A researcher randomly assigns 100 participants to two groups: one receives a new dietary supplement, and one receives a placebo. She measures the change in cholesterol levels after 12 weeks. Which type of study is this?

Show explanation

Correct answer: C. A randomized controlled experiment

Explanation

Random assignment to treatment and control groups defines a randomized controlled experiment. This design can support causal conclusions.

Question 4Medium

A poll of 1,000 randomly selected adults in a city found that 62% support a new transit policy, with a margin of error of ±3%. Which of the following is the best interpretation?

Show explanation

Correct answer: D. Between 59% and 65% of all adults in the city are estimated to support the policy

Explanation

Margin of error creates an interval estimate: 62% ± 3% = [59%, 65%]. This interval applies to the city's adult population (the sampling frame), not the country.

Question 5Medium

A nutrition scientist observes that people who eat breakfast daily tend to have lower body mass index (BMI) than those who skip breakfast. Based on this observational study, which conclusion is valid?

Show explanation

Correct answer: A. Eating breakfast is associated with lower BMI in this sample

Explanation

An observational study can only establish association, not causation. The conclusion is limited to the sample observed, using language like 'associated with.'

Question 6Medium

Two studies examine the effect of a new teaching method on test scores. Study 1 uses a random sample of 50 students; Study 2 uses 200 volunteers who sign up to participate. Which study produces results that can more reliably be generalized to all students?

Show explanation

Correct answer: B. Study 1, because random selection produces a representative sample

Explanation

Random sampling—not sample size—determines generalizability. Volunteers are self-selected and may differ systematically from all students, producing biased results regardless of sample size.

Question 7Medium

A study tests whether a new fertilizer increases crop yield. Farmers voluntarily apply the fertilizer to the fields they believe are most fertile. The results show higher yields in fertilized fields. What is the main flaw in this study design?

Show explanation

Correct answer: C. Confounding: more fertile fields were chosen for treatment, not randomly assigned

Explanation

Because farmers chose which fields to fertilize—selecting the best ones—soil fertility is a confounding variable. We cannot determine whether the fertilizer or the pre-existing soil quality caused the higher yields.

Question 8Hard

A school district surveys every student in three randomly selected schools about homework time and grade point average. The study finds a positive correlation. Which of the following conclusions is best supported?

Show explanation

Correct answer: D. Across the district, students who do more homework tend to have higher GPAs

Explanation

The schools were randomly selected from the district, so conclusions can be extended to the district—but not the country (no national random sample). The design is observational, so only association (not causation) can be inferred. Answer D accurately captures both limitations.

Question 9Hard

A researcher studying sleep patterns randomly assigns 200 college students to two groups. Group A sleeps 8 hours per night for 4 weeks; Group B sleeps 6 hours per night. At the end, Group A shows significantly higher scores on a cognitive test. Which conclusion is most justified?

Show explanation

Correct answer: A. For the 200 students in the experiment, the 8-hour sleep condition caused higher cognitive scores

Explanation

Random assignment supports a causal comparison between the two sleep conditions for the participants in this experiment. The stem does not say the participants were randomly sampled from a broader population, so it does not justify generalizing the result to all college students or all adults.

Question 10Hard

A researcher uses a convenience sample of 500 gym members to study exercise habits in the general adult population. The study finds that 85% of adults exercise at least 3 times per week. Why is this estimate likely unreliable for the general adult population?

Show explanation

Correct answer: C. Gym members are not a random or representative sample of all adults

Explanation

Gym members are systematically more likely to exercise than the general adult population—this is selection bias. No matter how large the sample is, a non-representative convenience sample cannot reliably estimate a population parameter.

Want more Statistical Inference and Study Design practice?

Open this skill in the question bank, then filter the available practice by difficulty and score band.

Open Question Bank

Common Mistakes to Avoid

Use this checklist to catch common traps while working through Statistical Inference and Study Design questions, then add your own patterns as you review missed answers.

  • !Choosing an answer that claims causation from an observational study
  • !Generalizing beyond the population from which the sample was drawn (e.g., extending results from one city to the nation)
  • !Confusing a larger sample with a more representative sample—size alone doesn't fix selection bias
  • !Treating correlation as implying directionality of cause (A causes B vs. B causes A)
  • !Selecting an answer that underclaims (refuses to generalize from a valid random sample) when one overclaims is the trap

Strategy Tips: Statistical Inference and Study Design

Before reading the answer choices, decide two things about the study: (1) was sampling random? (2) was assignment experimental or observational? Those two decisions rule out most wrong answers

Eliminate any answer choice that contains 'causes,' 'proves,' or 'demonstrates' unless the study is a randomized experiment

Match an inference to the sampled population and use association or causation language consistent with the study design

If the question asks what cannot be concluded, use your inference rules to identify the strongest overclaim in the choices—that is your answer

Other Problem Solving & Data Analysis Subtopics

10 worked questions for this skill

Master Statistical Inference and Study Design

Continue with this skill in the question bank, or test how it holds up under section timing before returning to the surrounding Problem Solving skills.

SAT® is a trademark registered by the College Board, which is not affiliated with, and does not endorse, this website.

Send feedback