Skip to content
SAT · Problem Solving and Data Analysis

Statistical inference

Statistical inference appears in 9 reviewed SAT questions and sits in Problem Solving and Data Analysis. Below are worked examples showing exactly how the exam tests it, with the answer marked and every distractor explained.

How Statistical inference is tested

Example 1Medium
A polling company surveys 500 residents of a city of 200,000 people. The survey finds that 62% support a new policy. Which of the following best describes this situation?
  1. AExactly 124,000 residents support the policy.
  2. BAll 200,000 residents support the policy by definition.
  3. CThe sample is too small to draw any conclusions.
  4. DIt is estimated that approximately 124,000 residents support the policy.Correct

Why D is right

62% of 200,000 = 124,000 is an estimate based on sampling. Estimates from samples are subject to sampling error. The data or units point to this result when the values are interpreted correctly.
Why the other 3 choices miss
A
Incorrect: Survey results provide estimates, not exact counts of a population; saying 'exactly' overstates certainty.
B
Incorrect: 62% (not 100%) of the sample supports the policy; surveys cannot claim universal support.
C
Incorrect: 500 is a reasonable sample size for a population of 200,000 if randomly selected.
Example 2Medium
The following table shows the results of a random sample: - Exercised this week: Yes = 140, No = 60 If the population is 3,000, how many people are estimated to NOT exercise this week?
  1. A1,200
  2. B900Correct
  3. C1,050
  4. D600

Why B is right

P(no exercise) = 60/200 = 30%. Estimated non-exercisers = 30% × 3,000 = 900. The data or units point to this result when the values are interpreted correctly.
Why the other 3 choices miss
A
Incorrect: 1,200 = 40% × 3,000; 40% is not the proportion who do not exercise in the sample.
C
Incorrect: 1,050 = 35% × 3,000; 35% is not the correct proportion.
D
Incorrect: 600 = 20% × 3,000; 20% is not the sample proportion for not exercising.
Example 3Medium
A researcher randomly selects 50 voters from a district of 10,000. Of those 50, 34 support Measure A. What is the best estimate of the number of voters in the district who support Measure A?
  1. A3,400
  2. B5,000
  3. C6,400
  4. D6,800Correct

Why D is right

Proportion supporting = 34/50 = 0.68. Estimate = 0.68 × 10,000 = 6,800. The data or units point to this result when the values are interpreted correctly.
Why the other 3 choices miss
A
Incorrect: 3,400 = 34 × 100, confusing the sample count with a proportion scaled differently.
B
Incorrect: 5,000 = 50% × 10,000; 50% is not the sample proportion (which is 68%).
C
Incorrect: 6,400 = 64% × 10,000; the proportion is 34/50 = 68%, not 64%.

Studied alongside

Other concepts from Problem Solving and Data Analysis.