SAT · Math · Problem-Solving and Data Analysis

Evaluating Statistical Claims on the Digital SAT: How to Answer These Questions

Whether you can judge what an observational study or experiment can conclude.

How to answer it

  1. Ask whether subjects were randomly assigned to groups.
  2. Ask whether they were randomly sampled, and from what population.
  3. Allow a causal claim only with random assignment, and only for people like those studied.

Worked example

An original question in the SAT format, already answered.

SAT MathWorked example

Researchers randomly assigned 300 volunteers to drink either green tea or water every day for a month. The green tea group reported better sleep.

Which conclusion is best supported?

  1. AGreen tea improves sleep for all adults.
  2. BGreen tea likely caused better sleep among people like these volunteers.
  3. CPeople who choose to drink tea sleep better.
  4. DNo conclusion is possible because the sample is too small.

Answer: B. Random assignment supports a causal conclusion, but the volunteers weren’t a random sample of all adults, so it applies to people like them.

The most common trap

Generalizing past the sample. Volunteers aren’t a random sample of everyone.

Practice it

Hourly Prep’s daily sets mix evaluating statistical claims with the rest of SAT Math, and send more of it your way when it’s a weak spot. See how the SAT prep plan works, or browse the other SAT skill guides.

Primary sources

Keep going

SAT · Math

Ratios, Rates, Proportions, and Units

Set up a proportion with matching units in the same positions, then solve. Convert units before you compare them.

Read the guide →
SAT · Math

Percentages

Percent change is the change divided by the original amount. Take the base from the starting value, not the new one.

Read the guide →
SAT · Math

One-Variable Data

Mean, median, mode and range respond differently to new values. An outlier moves the mean far more than the median.

Read the guide →
SAT · Math

Two-Variable Data

On a scatterplot, the slope of the line of best fit is the predicted change in y for each one-unit increase in x.

Read the guide →

Practice evaluating statistical claims in 15 minutes a day.

Short SAT sets on the test-day screen, with an explanation for every question.

Get the app