One Proportion Z-Test Calculator

Use this free online One Proportion Z-Test Calculator to determine whether an observed sample proportion (p̂) differs significantly from a hypothesized population benchmark (p₀). Enter your successes (x), sample size (n), and significance level (α) to instantly calculate the standardized z-score test statistic, exact one-tailed or two-tailed p-values, critical boundaries, and confidence intervals with automated normal approximation verification (np₀ ≥ 10 and n(1 − p₀) ≥ 10).

One Proportion Inputs
Test settings
Alternative hypothesis (H₁)
Significance level α

Live One Proportion Z-Distribution Curve

Drag the marker to test hypothetical proportion test statistics against the rejection region.

−4−3−2−10+1+2+3+4z = 1.60
Fine Adjust z:1.60
Rejection Zone (α = 0.05)p-value Area (0.1096)

Step-by-Step Mathematical Derivation

1

State Hypotheses

Compare population proportion p against claimed proportion p₀:

H₀: p = 0.5
H₁: p ≠ 0.5
2

Sample Proportion & Null SE

Under the null hypothesis, standard error uses the hypothesized p₀:

p̂ = x / n = 58 / 100 = 0.5800
SE₀ = √[p₀(1 − p₀) / n] = 0.0500
3

Compute Z-Statistic

Standardize the difference between observed and hypothesized proportion:

z = (p̂ − p₀) / SE₀
z = (0.5800 − 0.5) / 0.0500 = 1.6000
4

Evaluate Significance

Compare z against critical boundary and calculate p-value:

Critical cutoff z* = ±1.960 (α = 0.05)
P-Value = 0.1096
Consistent with H₀: Fail to reject H₀

What Is a One Proportion Z-Test Calculator?

A one proportion z-test calculator is an online inferential statistics tool that determines whether an observed sample proportion differs significantly from a hypothesized population proportion. The calculator evaluates binary categorical data (success or failure) by standardizing the observed sample proportion against the hypothesized value using the standard normal distribution (Z).

There are 4 main benefits of using a One Proportion Z-Test Calculator:

1. Instant Standard Error & Z-Score

Computes the exact null standard error SE₀ = √[p₀(1 − p₀)/n] and standard score without tedious square root calculations.

2. Automated Normality Verification

Verifies the success-failure condition (np₀ ≥ 10 and n(1 − p₀) ≥ 10) to confirm the normal approximation is mathematically valid.

3. Dual P-Value & Confidence Interval

Derives exact directional p-values alongside the Wilson or Wald confidence interval for the true population proportion.

4. Publication-Ready Reporting

Outputs a complete step-by-step hypothesis testing derivation suitable for academic research, business intelligence, and quality audits.

Statisticians and researchers use this test across 4 widespread domains:

  • Manufacturing process quality control: Testing whether the defect rate of finished goods meets quality standards.
  • Public opinion polling: Testing if voter support for a referendum differs significantly from a 50% majority benchmark.
  • Clinical trials: Comparing patient symptom relief rates against historical therapeutic baselines.
  • Conversion rate optimization: Validating whether website design changes yield higher customer checkout percentages.

How to Use the One Proportion Z-Test Calculator

To use the one proportion z-test calculator, enter your success count, total sample size, hypothesized proportion benchmark, select your alternative hypothesis, and click calculate.

  1. Enter Success Count (x): Input the observed count of successes or defect-free units recorded in your sample.
  2. Enter Sample Size (n): Input the total number of observations, independent trials, or respondents in your study.
  3. Enter Hypothesized Proportion (p₀): Input the baseline target proportion as a decimal between 0 and 1 (such as 0.50 for 50%).
  4. Select the Alternative Hypothesis:Choose a two-tailed test (p ≠ p₀), right-tailed test (p > p₀), or left-tailed test (p < p₀).
  5. Set the Significance Level (α): Select α = 0.05 for a standard 5% significance threshold, or α = 0.01 for stricter validation.
  6. Inspect Output: Review the calculated sample proportion (p̂), z-statistic, p-value, critical bounds, confidence interval, and interactive curve.

One Proportion Z-Test Formula and Standard Error

The test statistic formula standardizes the difference between the observed sample proportion and the hypothesized population proportion using the null standard error.

z = (p̂ − p₀) / √[p₀(1 − p₀) / n]

Formula Notation Breakdown:

  • p̂ = x / n: The observed sample proportion of successes.
  • p₀: The hypothesized population proportion under the null hypothesis.
  • n: The total sample size or number of independent trials.
  • SE₀ = √[p₀(1 − p₀) / n]: The standard error of the proportion under the null hypothesis.

Notice that the hypothesis test statistic uses p₀ in the standard error denominator, because hypothesis testing evaluates sampling behavior under the assumption that H₀ is true.

Confidence Interval Formulas:

Two-Sided Confidence Interval
CI = p̂ ± z_(α/2) · √[p̂(1 − p̂) / n]
One-Sided Confidence Bound
Upper: [0, p̂ + z_α · √[p̂(1 − p̂) / n]]
Lower: [p̂ − z_α · √[p̂(1 − p̂) / n], 1]

Key Assumptions & Sample Size Conditions

There are 3 mandatory assumptions required for a valid one proportion z-test:

1. Simple Random Sampling

The sample data must be gathered through random sampling from the target population to prevent selection bias.

2. Independence of Observations (10% Condition)

Each observation must be independent of other trials. The sample size n must not exceed 10% of the total population (n < 0.10N), if sampling without replacement from a finite population.

3. Success-Failure Condition (Normal Approximation Rule)

Both the expected number of successes and expected failures must be at least 10:

n · p₀ ≥ 10   and   n · (1 − p₀) ≥ 10

Effect Size (Cohen's h) and Power Analysis

Cohen's h quantifies the practical importance or effect size between the observed sample proportion and hypothesized proportion independently of sample size.

h = 2 · arcsin(√p̂) − 2 · arcsin(√p₀)
Small Effect Size|h| ≈ 0.20

Subtle variation requiring large sample sizes to detect.

Medium Effect Size|h| ≈ 0.50

Noticeable difference visible across moderate sample cohorts.

Large Effect Size|h| ≈ 0.80

Substantial real-world divergence between proportions.

Use this formula during prospective study planning to calculate the required sample size for desired statistical power:

n = [(z_(1−α/2) + z_(1−β))² · p₀(1 − p₀)] / (p_a − p₀)²

Where p_a is the minimum detectable alternative proportion, β is the Type II error rate, and 1 − β is the statistical power (typically 0.80 or 0.90).

Step-by-Step Worked Examples

Example 1: Manufacturing Process Quality Standards (Two-Tailed Test)

A quality engineering team evaluates whether a newly configured manufacturing line deviates from standard quality specifications. Standard baseline production mandates that 50% of parts (p₀ = 0.50) meet premium classification. In a random sample of n = 100 manufactured components, inspectors record x = 45 defect-free units (p̂ = 0.45). At α = 0.05, does the process quality differ significantly from standard?

Step 1: H₀: p = 0.50 vs. H_a: p ≠ 0.50 (Two-tailed test)
Step 2: α = 0.05 → Two-tailed critical z* = ±1.9600
Step 3: Check conditions: np₀ = 100(0.50) = 50 ≥ 10, n(1−p₀) = 50 ≥ 10 (Valid)
Step 4: SE₀ = √[0.50(1 − 0.50) / 100] = √0.0025 = 0.0500
Step 5: z = (0.45 − 0.50) / 0.05 = −1.0000
Step 6: Two-tailed p-value = 2 · (1 − Φ(1.00)) = 0.3173
Step 7: Fail to reject H₀ because p = 0.3173 > 0.05 and |z| = 1.00 < 1.960. Insufficient statistical evidence that process quality deviates from 50%.
Example 2: Pharmaceutical Clinical Trial Patient Relief (Two-Tailed Test)

A pharmaceutical manufacturer claims that 60% of patients (p₀ = 0.60) experience clinical relief from a new allergy formulation. In an independent clinical trial with n = 50 patients, researchers observe that x = 25 patients (50%) report positive relief. At α = 0.05, is this observed relief rate significantly different from the manufacturer's claimed 60%?

Step 1: H₀: p = 0.60 vs. H_a: p ≠ 0.60
Step 2: α = 0.05 → Two-tailed critical z* = ±1.9600
Step 3: Check conditions: np₀ = 50(0.60) = 30 ≥ 10, n(1−p₀) = 50(0.40) = 20 ≥ 10 (Valid)
Step 4: SE₀ = √[0.60(0.40) / 50] = √0.0048 = 0.06928
Step 5: z = (0.50 − 0.60) / 0.06928 = −1.4434
Step 6: Two-tailed p-value = 2 · (1 − Φ(1.4434)) = 0.14891
Step 7: 95% Confidence Interval = 0.50 ± 1.96 · √[0.50(0.50)/50] = [0.3614, 0.6386]
Step 8: Fail to reject H₀ because p = 0.1489 > 0.05 and the 95% CI includes 0.60. Continue monitoring patient relief, if larger patient cohorts become available.
Example 3: E-Commerce Conversion Rate Optimization (Right-Tailed Test)

A digital marketing analyst tests a redesigned checkout flow. Historical records establish that the baseline checkout conversion rate is 8% (p₀ = 0.08). The marketing team exposes n = 500 visitors to the redesign, yielding x = 55 conversions (p̂ = 11%). Using α = 0.05, test whether the redesign generates a significantly higher conversion rate.

Step 1: H₀: p ≤ 0.08 vs. H_a: p > 0.08 (Right-tailed test)
Step 2: α = 0.05 → Right-tailed critical z* = +1.6449
Step 3: Check conditions: np₀ = 500(0.08) = 40 ≥ 10, n(1−p₀) = 460 ≥ 10 (Valid)
Step 4: SE₀ = √[0.08(0.92) / 500] = √0.0001472 = 0.01213
Step 5: z = (0.11 − 0.08) / 0.01213 = 2.4727
Step 6: Right-tailed p-value = 1 − Φ(2.4727) = 0.00671
Step 7: Reject H₀ because p = 0.0067 ≤ 0.05 and z = 2.4727 > 1.6449. Adopt the redesigned checkout, if boosting customer conversion rate is the company's goal.

Alternative Statistical Tests When Assumptions Fail

Consider these alternative statistical procedures when your data characteristics deviate from standard test requirements:

Exact Binomial Test

Use an exact binomial test, if sample size conditions fail (np₀ < 10 or n(1 − p₀) < 10). It computes exact binomial probabilities without relying on normal distribution approximations.

Chi-Square Goodness-of-Fit

Use a chi-square goodness-of-fit test, if analyzing multinomial categorical variables with 3 or more outcome categories.

Two Proportion Z-Test

Use a two-proportion z-test, if comparing success rates between two distinct sample groups rather than comparing one group against a known standard.

How to Perform a One Proportion Z-Test in R, Python, & Excel

Execute one proportion tests using standard statistical software libraries and spreadsheet formulas:

R Implementation
# R: One Proportion Test
x <- 45    # Observed successes
n <- 100   # Total trials
p0 <- 0.50 # Hypothesized proportion

# Standard test without continuity correction
result <- prop.test(x = x, n = n, p = p0, alternative = "two.sided", correct = FALSE)
print(result)
Python Implementation (Statsmodels & Scipy)
from statsmodels.stats.proportion import proportions_ztest
import numpy as np
from scipy import stats

x, n, p0 = 45, 100, 0.50

# Compute test statistic and p-value
zstat, pval = proportions_ztest(count=x, nobs=n, value=p0, alternative='two-sided')
print(f"Z-Statistic: {zstat:.4f} | P-Value: {pval:.4f}")

# 95% Confidence Interval for proportion
p_hat = x / n
se_sample = np.sqrt(p_hat * (1 - p_hat) / n)
ci_low = p_hat - 1.96 * se_sample
ci_high = p_hat + 1.96 * se_sample
print(f"95% CI: [{ci_low:.4f}, {ci_high:.4f}]")
Microsoft Excel & Google Sheets
Sample Proportion (p̂):=x / n
Null Standard Error (SE₀):=SQRT(p0 * (1 - p0) / n)
Z Statistic:=(p_hat - p0) / SE0
Two-Tailed P-Value:=2 * (1 - NORM.S.DIST(ABS(z), TRUE))
Two-Tailed Critical Value (α = 0.05):=NORM.S.INV(1 - 0.05 / 2)

Frequently Asked Questions

What is the difference between one-tailed and two-tailed p-values?▼

A one-tailed p-value tests whether the population proportion is strictly greater than or strictly less than the hypothesized value, whereas a two-tailed p-value tests for difference in either direction. Use a one-tailed test, if you have a pre-specified directional research prediction. Use a two-tailed test, if deviation in either direction carries operational relevance.

How large should my sample size be for a one proportion z-test?▼

Both expected successes (n · p₀) and expected failures (n · (1 − p₀)) must be at least 10 for the normal approximation to be mathematically valid. For example, if p₀ = 0.10, you need a sample size of at least n = 100 trials. Use an exact binomial test, if your sample size fails to meet this threshold.

How do I interpret the confidence interval in relation to the hypothesis test?▼

The confidence interval provides a range of plausible values for the true population proportion at your chosen confidence level. Fail to reject the null hypothesis, if the hypothesized proportion p₀ falls inside the confidence interval. Reject the null hypothesis, if p₀ falls outside the calculated interval.

Why does the hypothesis test standard error differ from the confidence interval standard error?▼

The hypothesis test standard error uses the hypothesized parameter p₀ because it assumes H₀ is true, while the confidence interval uses the sample estimate p̂ to quantify empirical uncertainty around the observed sample. This difference reflects distinct statistical objectives between significance testing and parameter estimation.

What is continuity correction in a proportion z-test?▼

Continuity correction adjusts the discrete binomial count by 0.5 when approximating it with the continuous normal distribution. Apply continuity correction, if your sample size is close to the minimum threshold of 10 and you require conservative p-value estimates.

Can I use the one proportion z-test if my hypothesized proportion is 0 or 1?▼

No, you cannot use a z-test when p₀ equals 0 or 1 because the standard error formula would divide by zero or result in zero variance. Use exact probability methods or Poisson modeling, if assessing rare deterministic events.

William Davis

Created by William Davis

Last updated: October 24, 2026

Add to Google Preferred