The pooled (combined) proportion, p̂c = (n₁p̂₁ + n₂p̂₂)/(n₁ + n₂), is the single best estimate of the common population proportion when the null hypothesis assumes p₁ = p₂; it's used in the standard error of the two-proportion z-test and to verify the normal/Large Counts condition.
The pooled proportion (the CED also calls it the "combined proportion," written p̂c) is what you get when you dump both samples into one pile and find the overall proportion of successes. The formula p̂c = (n₁p̂₁ + n₂p̂₂)/(n₁ + n₂) looks fancy, but the numerator is just total successes from both samples and the denominator is total people sampled.
Why combine them at all? Because a two-proportion z-test starts by assuming the null hypothesis H₀: p₁ = p₂ is true. If the two population proportions really are equal, then both samples are estimating the same number, and your best estimate of that shared number uses all the data at once. That's p̂c. You then use it in two places. First, you plug it into the standard error inside the test statistic z = (p̂₁ - p̂₂) / (√(p̂c(1-p̂c)) √(1/n₁ + 1/n₂)). Second, you use it to check that the sampling distribution of p̂₁ - p̂₂ is approximately normal by confirming the expected counts of successes and failures in each sample (using p̂c) are large enough.
Pooled proportion lives in Unit 6 (Inference for Categorical Data: Proportions), specifically Topics 6.10 and 6.11. It directly supports AP Stats 6.10.C (verifying conditions for a two-proportion z-test, where VAR-6.J.1 says to define p̂c when checking that the sampling distribution is approximately normal) and AP Stats 6.11.A (calculating the test statistic, whose standard error formula is built on p̂c). It also connects to AP Stats 6.11.B, since interpreting the p-value means recognizing it was computed assuming the proportions are equal, which is exactly the assumption that justifies pooling. Per the CED's clarifying statement, the test statistic formula isn't printed on the AP formula sheet, so you need to be able to build it yourself, and knowing when and why to pool is the part that trips people up.
Keep studying AP® Statistics Unit 3
Null Hypothesis (Unit 6)
Pooling only makes sense because H₀: p₁ = p₂ says both populations share one proportion. The pooled proportion is your estimate of that one shared value, so the whole concept lives or dies on the null hypothesis assumption.
Large Counts Condition (Units 5-6)
For a two-proportion test, you check large counts with p̂c, not the individual sample proportions. You need n₁p̂c, n₁(1-p̂c), n₂p̂c, and n₂(1-p̂c) all large enough for the sampling distribution of p̂₁ - p̂₂ to be approximately normal.
Confidence Interval for a Difference in Proportions (Unit 6)
A confidence interval makes no assumption that p₁ = p₂, so there is nothing to pool. It uses each sample's own p̂ in the standard error. This is the single most-tested distinction between the test and the interval.
Difference in Two Population Proportions (Unit 6)
The pooled proportion is a tool for studying the sampling distribution of p̂₁ - p̂₂. Under the null, that distribution centers at 0, and p̂c gives you its standard error and its shape check.
Multiple-choice questions love two angles on this term. One gives you raw counts (like 210 of 400 supporting a policy in Region 1 and 180 of 300 in Region 2) and asks you to compute p̂c or the test statistic built from it. The other asks the conceptual "why" question, like which statement best explains why the pooled proportion replaces the individual sample proportions in the standard error. The answer is always the same idea, that the test assumes H₀: p₁ = p₂ is true. On the FRQ side, a full significance test for two proportions requires you to show the pooled proportion when checking the normality condition and when calculating z, and slightly different p-values between two analysts (0.047 vs. 0.052) often trace back to one using pooled and the other using unpooled standard error. Remember, the test statistic formula isn't on the formula sheet, so practice constructing it.
Use pooled p̂c only in a significance test, because the null hypothesis assumes the two proportions are equal and pooling estimates that one shared value. Use the separate p̂₁ and p̂₂ in a confidence interval for p₁ - p₂, because an interval makes no equality assumption. A quick rule of thumb is tests pool, intervals don't.
The pooled proportion is total successes from both samples divided by total sample size, p̂c = (n₁p̂₁ + n₂p̂₂)/(n₁ + n₂).
You pool because the null hypothesis H₀: p₁ = p₂ assumes one common proportion, and combining both samples gives the best estimate of it.
The pooled proportion shows up in two places in a two-proportion z-test, inside the standard error of the test statistic and in the large counts check for normality.
Confidence intervals for p₁ - p₂ never use the pooled proportion, because intervals don't assume the proportions are equal.
The two-proportion test statistic formula isn't printed on the AP formula sheet, so you have to build it from the general test statistic structure and remember to pool.
It's the combined estimate of a common population proportion, p̂c = (n₁p̂₁ + n₂p̂₂)/(n₁ + n₂), which is just total successes over total sample size from both groups. It's used in the two-proportion z-test from Topics 6.10 and 6.11.
No. Pooling is only justified when you assume p₁ = p₂, which is the null hypothesis of a significance test. A confidence interval for p₁ - p₂ makes no such assumption, so it uses the separate sample proportions p̂₁ and p̂₂ in the standard error.
Because the p-value is computed assuming H₀: p₁ = p₂ is true. If both populations share one proportion, both samples are estimating the same number, and combining all the data into p̂c gives the most accurate estimate of it.
Use the pooled proportion, not the individual ones. Check that n₁p̂c, n₁(1-p̂c), n₂p̂c, and n₂(1-p̂c) are all at least 10, which confirms the sampling distribution of p̂₁ - p̂₂ is approximately normal.
Not directly. The CED notes that the two-proportion test statistic formula doesn't appear explicitly on the formula sheet, but you can construct it from the general test statistic and standard error formulas that are provided. Knowing that p̂c is total successes over total n makes it easy to rebuild.
Connect this key term to the AP exam workflow: review the course, practice questions, and check related study tools.
Review units, study guides, and course resources.
Check this vocabulary in multiple-choice context.
Apply key concepts in written AP responses.
Estimate the exam score you are working toward.
Review the highest-yield facts before practice.
Put the full course together before test day.