Hypothesis Tests About a Population Proportion Large Samples the Critical Value Approach

Statistics • Hypothesis Tests About the Mean and Proportion

Written by STEM Calculators Team Published December 24, 2025 Updated February 24, 2026

Hypothesis Tests About a Population Proportion: Large Samples

Use the critical-value approach for a one-sample proportion z test: choose H₀ / H₁, pick α, find the critical z value(s), compute the observed z, and decide whether to reject H₀.

Test setup

Alternative direction (tail)

α (significance level)

Visualization shading

The null hypothesis must include equality (e.g., p = p₀, p ≥ p₀, or p ≤ p₀). The alternative hypothesis determines whether the test is two-tailed, left-tailed, or right-tailed.

Null value and sample information

Null value p₀ (from H₀)

Basic probability constraints: 0 ≤ p ≤ 1 and q = 1 − p.

How will you provide sample values?

Sample size n

Successes x

Then p̂ = x / n.

Assumed true p (optional, for β and power)

If provided, the calculator estimates β and power (1 − β).

Quick reference: tails and critical values

Tail type	H₁ sign	Critical z value(s)	Reject H₀ when…
Two-tailed	p ≠ p₀	±z_α/2	\|z\| ≥ z_α/2
Left-tailed	p < p₀	z_α (negative)	z ≤ z_α
Right-tailed	p > p₀	z_1−α (positive)	z ≥ z_1−α

The test statistic uses the null value p₀ in the standard error: SE₀ = √(p₀(1−p₀

Ready

Rate this calculator

0.0 /5 (0 ratings)

Be the first to rate.

Your rating

Name (optional) Review (optional)

You can update your rating any time.

Hypothesis Tests About a Population Proportion (Large Samples): The Critical-Value Approach

Use this method to test a claim about a population proportion p when the sample is large enough for the normal approximation. The calculator computes the test statistic z, selects the correct critical value(s), and applies the reject / do-not-reject rule.

Notation

Population proportion: \(p\)
Null value: \(p_{0}\) (the value stated in \(H_{0}\))
Sample size: \(n\), successes: \(x\)
Sample proportion: \(\hat{p}=\dfrac{x}{n}\)
Complement under the null: \(q_{0}=1-p_{0}\)
Significance level: \(\alpha\)

Large-sample (normal) condition

Check the expected counts using the null value:

\[ n \cdot p_{0} \ge 5 \quad \text{and} \quad n \cdot (1-p_{0}) \ge 5. \]

If these fail, the z approximation may be unreliable and an exact/binomial method is preferred.

Critical-value procedure (z test for \(p\))

State \(H_{0}\) and \(H_{1}\) (the alternative determines the tail):

\[ \text{Two-tailed: } H_{0}:p=p_{0}\ \ \text{vs.}\ \ H_{1}:p \ne p_{0} \] \[ \text{Left-tailed: } H_{0}:p=p_{0}\ \ \text{vs.}\ \ H_{1}:p < p_{0} \] \[ \text{Right-tailed: } H_{0}:p=p_{0}\ \ \text{vs.}\ \ H_{1}:p > p_{0} \]
Compute the standard error under \(H_{0}\):

\[ \mathrm{SE}_{0}=\sqrt{\frac{p_{0}(1-p_{0})}{n}} \;=\; \sqrt{\frac{p_{0}\cdot q_{0}}{n}}. \]
Compute the test statistic:

\[ z=\frac{\hat{p}-p_{0}}{\mathrm{SE}_{0}}. \]
Find critical value(s) from \(\alpha\):

\[ \text{Two-tailed: critical values } \pm z_{\alpha/2} \] \[ \text{Left-tailed: critical value } -z_{\alpha} \] \[ \text{Right-tailed: critical value } z_{\alpha} \]
Decision rule:

\[ \text{Two-tailed: reject } H_{0}\ \text{if}\ |z|\ge z_{\alpha/2} \] \[ \text{Left-tailed: reject } H_{0}\ \text{if}\ z\le -z_{\alpha} \] \[ \text{Right-tailed: reject } H_{0}\ \text{if}\ z\ge z_{\alpha} \]

What the calculator can take as input

Summary: enter n and x (it computes \(\hat{p}\)).
Paste / CSV: paste a column of 0/1 values (or yes/no, true/false) or upload a CSV, then it computes \(x\), \(n\), and \(\hat{p}\).

Frequently Asked Questions

What is the critical value approach for a hypothesis test about a population proportion?

The critical value approach compares the observed z test statistic to cutoff value(s) determined by alpha and the tail direction. You reject H0 if z falls in the rejection region defined by the critical z value(s).

How is the z test statistic for a population proportion calculated?

First compute p-hat = x/n. Under H0, the standard error is SE0 = sqrt(p0(1 - p0)/n), and the test statistic is z = (p-hat - p0)/SE0.

How do you find the critical z values for two-tailed and one-tailed tests?

For a two-tailed test, the critical values are plus/minus z(alpha/2). For a left-tailed test the cutoff is negative and you reject when z is at or below the left critical value; for a right-tailed test the cutoff is positive and you reject when z is at or above the right critical value.

When is the large-sample normal approximation valid for a proportion z test?

A common guideline is that expected counts under the null are at least 5: n x p0 >= 5 and n x (1 - p0) >= 5. If these are not met, the z approximation can be unreliable.

What do beta and power mean in this calculator?

Beta is the probability of not rejecting H0 when the true proportion differs from p0 in the direction of H1 (a Type II error). Power is 1 - beta and represents the probability of correctly rejecting H0 for the assumed true proportion.

Hypothesis Tests About a Population Proportion: Large Samples

Calculation steps

Rate this calculator

Frequently Asked Questions

Related calculators