This pooled proportion calculator combines two or more samples into one proportion, p̄. Enter the number of successes for each group with its group size, or two percents. It returns p̄ as an exact fraction, a decimal and a percent. For two groups it also gives the pooled standard error. The z statistic is the one the two proportion z test uses.
What Is a Pooled Proportion?
A pooled proportion is the share of successes when two samples are treated as one. It counts favorable cases considering both samples together: total successes over total observations. The symbol is p̄ (p-bar), sometimes written as a population proportion with hat, p̂, with no subscript.
Pooling follows from the null hypothesis. It says the groups share one common population proportion under H₀. If that is true, both samples estimate the same number, and combining them gives a better estimate than either alone.
Pooled Proportion Formula
Formula
p̄ = (x₁ + x₂) ÷ (n₁ + n₂) x = the number of successes, n = group size.
For many groups the same idea becomes p̄ = Σx ÷ Σn. It equals the weighted average of the group proportions, each weighted by its n: p̄ = (n₁p̂₁ + n₂p̂₂) ÷ (n₁ + n₂).
Formula
SE pooled = √( p̄(1 − p̄)(1/n₁ + 1/n₂) ) The pooled standard error of p̂₁ − p̂₂.
Formula
z = (p̂₁ − p̂₂) ÷ SE pooled Worked Example
An A/B test converts 80 of 200 visitors on page A and 60 of 200 on page B.
- p̄ = (80 + 60) ÷ (200 + 200) = 140/400 = 7/20 = 0.35.
- Pooled SE = √(0.35 × 0.65 × (1/200 + 1/200)) ≈ 0.0477.
- z = (0.40 − 0.30) ÷ 0.0477 ≈ 2.097, the same z our two proportion z test reports.
Here is a pooled proportion example with three groups: 12/40, 30/90 and 18/70. The result is 60/200 = 0.30, even though the group rates are 30%, 33.3% and 25.7%.
When to Pool and When Not To
- Pool for a two sample proportion test of H₀: p₁ = p₂, because the test assumes the proportions are equal.
- Do not pool for a confidence interval for p₁ − p₂: that interval estimates how different the proportions are, so each group keeps its own variance.
- Pool only independent samples measured the same way; paired data need other methods.
- Check successes and failures: each group should have at least about 10 of each for the normal approximation.
For the unpooled interval, see our two proportion confidence interval calculator. That page explains why proportions confidence interval methods keep the groups apart.
Pooled vs Unpooled Standard Error
In the example, the pooled SE is 0.0477 and the unpooled SE is 0.0474. The two are close when the sample sizes are similar. They drift apart when one group is much larger and the proportions differ, which is when the choice matters most.
Pooled Proportion Calculator Uses
Students use it to check their pooled proportion calculation by hand. It is the first step of the z test for two groups. Analysts use it to combine batches, shifts or survey waves into one rate. It is also the first step of a proportion z test calculator, before the p-value.
Some call it a pooled proportion estimator. Others call it a combined proportion calculator. A third name is aggregate proportion tool.
The input data fields take whole counts, or percents with sample sizes. The pooling algorithm adds the counts as exact integers, so p̄ appears as a reduced fraction. The output display shows p̄ as a fraction, a decimal and a percent.
Notation in Plain Text
Software and forums often type the symbols without subscripts:
- H0: the null hypothesis that the proportions are equal.
- H1: the alternative hypothesis.
- x1: successes in group 1.
- x2: successes in group 2.
- n1: sample size of group 1.
- n2: sample size of group 2.
- p1: the true proportion in group 1.
- p2: the true proportion in group 2.
- p̂1: the sample proportion of group 1.
- p̂2: the sample proportion of group 2.
Pooled Proportion in Statistics
A pooled proportion definition is short: all successes divided by all trials. How to calculate pooled proportion values is just as short: add, then divide. To calculate pooled proportion for two samples, use the formula above.
This pooled proportion equation also has a name in textbooks. Many call p̄ the pooled sample proportion. It is a pooled estimate of the shared rate. In pooled proportion statistics, the symbol p̄ always means this weighted proportion.
A pooled proportion test is another name for the two-group z test. A pooled proportion z-test puts p̄ into its SE. A pooled proportion hypothesis test always starts from H₀: p₁ = p₂. Its pooled proportion standard error is the √(p̄(1 − p̄)(1/n₁ + 1/n₂)) line in the steps.
A pooled proportion two samples problem needs four numbers. A pooled proportion estimate is then a single rate for the pooled sample. With pooled data from many groups, the third tab does the sums. A pooled proportion confidence interval for p̄ itself is the Wilson interval of Σx out of Σn.
Pooling is the proportion version of pooled variance in a t test. The proportion difference p̂₁ − p̂₂ is what the z test examines. A proportion hypothesis test of this kind is common in A/B testing.
This page is a pooled proportion online calculator, so nothing is installed. It is a statistical calculator for one formula. It works as an online calculator on any phone. You can also calculate proportion values for each group from its own counts.
Related Terms
A z-test compares the z statistic with the standard normal distribution. The binomial distribution describes the counts, and the normal approximation needs a large sample size. The significance level, alpha, is usually 0.05. A continuity correction can be added for small counts. Effect size, such as Cohen’s h, describes how big the difference is. Each group’s sample proportion estimates its own true rate. The standard deviation of the difference uses p̄.
A Two sample proportion test calculator works with pooled proportions inside the test. With only one group, no pooling is needed. That is the job of a One sample proportion calculator. Two proportions confidence interval methods and a single Proportions confidence interval do not pool either. Every confidence interval for a difference keeps the groups apart.
Independent samples are required. The steps on this page are a step-by-step solution you can copy into homework. For a single group, use our population proportion calculator. It will estimate the population proportion with an interval.
Pooled Proportion Calculator FAQ
What is a pooled proportion?
The combined share of successes from two or more samples: total successes over total sample size.
How do you calculate a pooled proportion?
Add the successes and divide by the total sample size: (80 + 60) ÷ (200 + 200) = 0.35.
Why is the pooled proportion used in the two proportion z test?
Because the null hypothesis assumes both groups share one proportion, so the best estimate of it combines both samples.
Is the pooled proportion the average of the two proportions?
Only when the sample sizes are equal; otherwise it is the average weighted by sample size.
Should I pool for a confidence interval?
No, intervals for p₁ − p₂ use the unpooled standard error.
Is this the same as a pooled proportion in meta-analysis?
No. A meta-analytic proportion calculator combines published studies with weights. A fixed-effect pooled proportion calculator weights each study by its precision. A random-effects pooled proportion calculator also allows for differences between studies. Some first transform the proportions, as a logit-transformed pooled proportion calculator does. A Freeman-Tukey double arcsine pooled proportion calculator uses another transform.
References
Related Calculators
- Two Proportion Z Test Calculator: compare two proportions with a z test
- Two Proportion Confidence Interval Calculator: interval for the difference of two proportions