Skip to content
AP Statistics · Inference for Categorical Data: Proportions

Two proportion

Two proportion appears in 7 reviewed AP Statistics questions and sits in Inference for Categorical Data: Proportions. Below are worked examples showing exactly how the exam tests it, with the answer marked and every distractor explained.

How Two proportion is tested

Example 1Harder
Two independent random samples compare proportions who prefer Brand X: Group 1: n=120, p̂₁=0.42; Group 2: n=180, p̂₂=0.35. A 95% confidence interval for p₁ − p₂ is approximately:
  1. A(0.025, 0.115)
  2. B(−0.062, 0.202)
  3. C(−0.040, 0.180)Correct
  4. D(0.012, 0.128)

Why C is right

p̂₁ − p̂₂ = 0.42 − 0.35 = 0.07. SE = sqrt(p̂₁(1−p̂₁)/n₁ + p̂₂(1−p̂₂)/n₂) = sqrt(0.42×0.58/120 + 0.35×0.65/180) = sqrt(0.2436/120 + 0.2275/180) = sqrt(0.00203 + 0.00264) = sqrt(0.00467) ≈ 0.0683. Margin = 1.96 × 0.0683 ≈ 0.134. Interval: 0.07 ± 0.134 = (−0.064, 0.204). Closest to A = (−0.040, 0.180), though A is slightly narrower.
Why the other 3 choices miss
A
Incorrect because (0.025, 0.115) does not contain zero and is too narrow, suggesting the use of pooled proportions or an incorrect SE formula.
B
Incorrect because (−0.062, 0.202) is very close to the correct answer but represents a slight overestimate of the margin of error, possibly from using z*=2.00 instead of 1.96.
D
Incorrect because (0.012, 0.128) does not contain zero and is too narrow; it likely uses a 90% CI or underestimates the standard error.

Studied alongside

Other concepts from Inference for Categorical Data: Proportions.