Simple Sample Size Calculation Formula Calculator
Estimate the minimum sample size needed for surveys and proportion studies using the standard formula for confidence level, margin of error, and expected proportion. Optional finite population correction is included.
Higher confidence requires a larger sample.
Common values are 3%, 5%, or 10%.
Use 50% if you do not know the expected proportion.
Leave blank for a large or unknown population.
This calculator uses the classic formula n = Z² × p × (1-p) / e² and applies finite population correction when population size is entered.
Your results will appear here
Enter your research assumptions, then click the calculate button to see the required sample size, infinite population estimate, and finite population correction.
Expert Guide to the Simple Sample Size Calculation Formula
The simple sample size calculation formula is one of the most widely used tools in survey design, market research, social science, public health, education studies, and quality control. When researchers want to estimate a percentage in a population, such as the share of customers who prefer a product, the percentage of residents who support a policy, or the proportion of patients who respond to a treatment, they need to know how many observations are enough to produce a useful estimate. That is exactly where the sample size formula becomes essential.
At its core, the idea is straightforward. A larger sample gives a more precise estimate, but collecting more data costs time and money. The simple sample size calculation formula balances precision and practicality. It helps determine the minimum number of responses needed to achieve a desired confidence level and margin of error under a proportion-based study framework.
The classic formula for an infinite or very large population is:
n = (Z² × p × (1-p)) / e²
In this equation, Z is the z-score associated with the confidence level, p is the estimated proportion, and e is the allowable margin of error expressed as a decimal. For example, 5% becomes 0.05.
What each part of the formula means
To use the formula correctly, it is important to understand every term:
- Sample size (n): the number of completed responses or observations needed.
- Confidence level: how certain you want to be that the true population value falls inside your interval estimate. Common choices are 90%, 95%, and 99%.
- Z-score: the statistical multiplier tied to the selected confidence level. Typical values are 1.645 for 90%, 1.96 for 95%, and 2.576 for 99%.
- Estimated proportion (p): the expected share of the population with the attribute of interest. If you do not know it, using 50% or 0.50 is standard because it produces the most conservative sample size.
- Margin of error (e): the amount of estimation error you are willing to tolerate. Smaller error means a larger sample.
Suppose you want to estimate the percentage of website visitors who would subscribe to a paid plan. If you do not know the expected conversion share, you might use p = 0.50, choose a 95% confidence level with Z = 1.96, and a 5% margin of error with e = 0.05. Plugging those values into the formula gives:
n = (1.96² × 0.50 × 0.50) / 0.05² = 384.16
Because sample size must be rounded up, you would need 385 respondents.
Why 50% is often used in practice
Researchers frequently use 50% when they have no prior estimate for p. This is not arbitrary. The expression p × (1-p) reaches its maximum value when p = 0.50. That means the formula produces the largest required sample at 50%, giving a conservative design. If the true proportion turns out to be 20% or 80%, the minimum sample needed for the same confidence level and margin of error would actually be smaller.
For planning purposes, this conservative assumption can be useful because it reduces the risk of underpowering a survey. In client work, public policy surveys, and academic proposals, it is common to justify a sample using 50% unless a pilot study or prior literature suggests a more accurate estimate.
Finite population correction and when it matters
The basic formula assumes an effectively unlimited population. But what if your target population is not large? For example, perhaps there are only 1,000 students in a department, 2,500 employees in a company, or 700 households in a small community. In that case, the infinite population estimate can overstate the number of observations you actually need.
This is where the finite population correction is applied:
n-corrected = n / (1 + ((n – 1) / N))
Here, N is the total population size and n is the initial sample estimate from the simple formula.
For example, if your uncorrected sample size is 385 but the total population is only 1,000, the corrected sample becomes approximately 278. This can significantly reduce fieldwork effort without sacrificing the intended statistical precision.
Comparison table: common sample sizes for large populations
The table below shows approximate sample sizes for large populations using the classic formula and a conservative proportion of 50%.
| Confidence Level | Z-score | Margin of Error | Estimated Proportion | Approximate Sample Size |
|---|---|---|---|---|
| 90% | 1.645 | 5% | 50% | 271 |
| 95% | 1.96 | 5% | 50% | 385 |
| 99% | 2.576 | 5% | 50% | 664 |
| 95% | 1.96 | 3% | 50% | 1,067 |
| 95% | 1.96 | 2% | 50% | 2,401 |
These figures illustrate one of the most important practical lessons in sample size planning: reducing the margin of error has a dramatic impact. Moving from a 5% margin of error to a 3% margin does not just increase the sample a little. It can nearly triple it.
Finite population examples with real numeric outputs
The next table shows how finite population correction changes the result for a 95% confidence level, 5% margin of error, and 50% estimated proportion. The infinite population sample is 385 before correction.
| Total Population Size | Initial Infinite Population Sample | Corrected Sample Size | Approximate Reduction |
|---|---|---|---|
| 500 | 385 | 218 | 43% |
| 1,000 | 385 | 278 | 28% |
| 5,000 | 385 | 357 | 7% |
| 10,000 | 385 | 370 | 4% |
As the population grows, the corrected sample moves closer to the infinite population result. This is why many national surveys with very large target populations still rely on the classic uncorrected benchmark.
Step by step: how to calculate sample size manually
- Choose your confidence level, such as 95%.
- Find the related z-score, such as 1.96 for 95% confidence.
- Select a margin of error, such as 5%, and convert it to decimal form: 0.05.
- Estimate the expected proportion p. If unknown, use 0.50.
- Apply the formula n = Z² × p × (1-p) / e².
- Round up to the next whole number.
- If your population is small or known, apply finite population correction.
This process works well for a broad range of simple survey studies. It is especially suitable when the main outcome is binary or proportion-based, such as yes or no responses, support or opposition, adoption or non-adoption, or success or failure.
Common mistakes to avoid
- Using percentages instead of decimals inside the formula: 5% must be entered as 0.05, not 5.
- Rounding down sample size: sample size should almost always be rounded up.
- Ignoring nonresponse: if you need 385 completed responses and expect a 50% response rate, you may need to contact roughly 770 people.
- Confusing confidence level with response rate: these are entirely different concepts.
- Assuming the formula handles complex designs: cluster sampling, stratified weighting, and experimental designs may require extra adjustments or design effects.
Adjusting for nonresponse and real field conditions
The simple sample size formula gives the number of completed observations required, not the number of invitations you should send. Real world studies nearly always experience drop-off, ineligibility, missing data, or refusals. If you estimate a 60% response rate, divide your required complete sample by 0.60 to determine how many people you should attempt to recruit.
For example, if your required completed sample is 385 and you expect a 60% response rate, then your outreach target is 385 / 0.60 = 641.67. Rounded up, you should invite at least 642 potential respondents.
When the simple formula is appropriate
This formula is ideal for:
- Opinion polls and customer satisfaction surveys
- Community needs assessments
- Website user research involving binary outcomes
- Proportion-based health, education, and social science studies
- Quality assurance checks when estimating defect or compliance rates
However, if you are comparing means, testing differences between groups, building regression models, or conducting a randomized controlled trial, you may need a different sample size framework. In those cases, power analysis is often more appropriate than the simple proportion formula.
How authoritative sources frame sample size decisions
Government and university statistical resources consistently emphasize the relationship among confidence, precision, variability, and population size. The U.S. Census Bureau provides extensive methodological resources on survey estimation and margins of error. The Centers for Disease Control and Prevention publishes epidemiologic guidance and field methods where sample assumptions influence disease surveillance and prevalence studies. Academic references from institutions such as UC Berkeley Statistics also explain core sampling logic and inferential principles used in this calculator.
Best practices for using a sample size calculator
- Start with a realistic study objective. Define exactly what proportion you want to estimate.
- Use 95% confidence and 5% margin of error as a reasonable baseline if no stricter requirement exists.
- Choose p = 50% when prior data is unavailable.
- Apply finite population correction if the target population is modest in size.
- Add a buffer for nonresponse, screen-outs, and unusable records.
- Document your assumptions in proposals, methods sections, and stakeholder reports.
Final takeaway
The simple sample size calculation formula remains one of the most practical and reliable planning tools in applied research. It converts abstract goals like confidence and precision into a concrete sample target you can budget, recruit, and defend. For many proportion-based studies, the formula n = Z² × p × (1-p) / e² provides a clear foundation, while finite population correction improves realism when the total population is limited.
If you remember only one principle, make it this: every gain in precision costs more data. A narrower margin of error or higher confidence level quickly expands sample requirements. By understanding this tradeoff and using a calculator like the one above, you can make informed, statistically sound decisions before your study begins.