How to solve Evaluating Statistical Claims questions on the digital SAT
Ask two questions about any study. Was the sample chosen at random from a larger group? If so, you can generalize to that group. Were subjects randomly assigned to treatments? If so, you can talk about cause and effect. A study with neither can only describe the people in it.
The method
Step 1: Check selection. Random selection lets you generalize, but only to the group selected from.
Step 2: Check assignment. Random assignment to treatments lets you claim cause and effect.
Step 3: Name the scope. Which group do the results reach? Find the choice that stays inside it.
Step 4: Remove overreach. Cross out choices with cause-and-effect words or larger groups that the study does not support.
How to spot this question
The question describes a survey or study, then asks which conclusion is supported or how to improve it. The numbers matter less than how the people were chosen and treated.
Example 1
Try each one yourself before you read on. These practice questions were written for this guide, and their names and numbers are made up.
Researchers randomly selected 500 adults from a city's list of residents. They then randomly assigned half to a daily walking program and half to no program. After 6 months, the walking group had lower average blood pressure. Which conclusion is best supported?
A) The walking program lowers blood pressure for people in every city.
B) The walking program likely lowers blood pressure, on average, for adults in this city.
C) Lower blood pressure causes adults in this city to walk more.
D) No conclusion can be made, because only 500 adults took part.
Working through example 1
Selection was random from the city, so results can reach adults in this city.
Assignment to the program was random, so a cause-and-effect claim is fair.
Both checks pass, but the scope stops at this city. The conclusion should say so.
Example 1: the answer is B
A: The sample came from one city, so the claim cannot reach every city.
B: Random selection supports this city, and random assignment supports the word "lowers."
C: This reverses the direction. The program was assigned, then blood pressure was measured.
D: A random sample of 500 is enough to support a conclusion. The size alone does not rule one out.
Example 2
A website asks visitors to report their sleep and their grades. Those who report more sleep also report higher grades. Which conclusion is best supported?
A) Getting more sleep causes students to earn higher grades.
B) Earning higher grades causes students to sleep more.
C) Among the visitors who responded, more reported sleep goes with higher reported grades.
D) Students across the country who sleep more earn higher grades.
Working through example 2
Visitors chose to respond, so selection was not random. The results stay with the people who answered.
Nobody was assigned an amount of sleep, so this study cannot show cause and effect.
Only a description of the people who responded is left.
Example 2: the answer is C
A: Nobody was randomly assigned to sleep more, so this study cannot show cause.
B: This makes the same mistake in the other direction.
C: It describes only the people who responded, and it claims no cause.
D: People who chose to respond do not stand for all students, so this reaches too far.
Traps to watch for
Cause from a survey: Without random assignment, a link between two things does not show that one causes the other.
Volunteers as a random sample: People who choose to respond can differ from people who do not.
Too large a group: Results reach only the group that was randomly sampled, such as one school, not all students.
Quick answers
What is the difference between random selection and random assignment?
Random selection decides who is in the study and supports generalizing. Random assignment decides who gets which treatment and supports cause and effect.
Can an observational study prove cause?
No. It can show a link, but something else could explain it. A randomized experiment is what supports a cause-and-effect claim.
What is sampling bias?
It is when the way people are chosen makes some more likely to be included, so the sample does not match the whole group.
Your next step
Now try it on questions you have not seen. The free diagnostic shows which skills to study first, and the free sampler lets you try questions with explanations. A study plan opens all 4 chapter practice questions on Evaluating Statistical Claims, with every answer choice explained.
Official facts about this section, such as the number of questions in each content area and the order of questions in a module, were checked against the official pages below on 8 November 2026. The methods and the examples are our own teaching. Confirm details on the College Board website before test day.
Skill by Skill Prep is independent. SAT® and AP® are trademarks registered by the College Board, which is not affiliated with, and does not endorse, this website or book. PSAT/NMSQT® is a registered trademark of the College Board and the National Merit Scholarship Corporation, which are not affiliated with, and do not endorse, this site. Bluebook™ is a trademark of the College Board. ACT® is a trademark of ACT Education Corp., which is not affiliated with and does not endorse this website or book. Desmos is a trademark of Desmos Studio PBC. Practice results describe these questions, not official scores or admission predictions.