Sample Size Calculator

What Is a Sample Size Calculator?

A sample size calculator estimates how many people, responses, observations, or data points you need to survey to estimate a population proportion with a chosen level of confidence and precision.

Instead of guessing how many responses are enough, you can calculate sample size using your confidence level, margin of error, population proportion, and population size.

This online sample size calculator can also work in reverse. If you already know your sample size, you can use the Find Margin of Error option to estimate the margin of error for your survey or study.

For example, with a 95% confidence level, 5% margin of error, 50% population proportion, and a large population, the required sample size is 385.

How to Use the Sample Size Calculator

The calculator provides two options: Find Sample Size and Find Margin of Error.

How to Find Sample Size

Use this option when you know the precision you want but need to determine how many responses to collect.

1. Select Find Sample Size.

2. Choose your Confidence Level.

3. Enter your desired Margin of Error as a percentage.

4. Enter the expected Population Proportion. Use 50% when you do not have a reliable estimate.

5. Enter the Population Size if you know the total size of a finite population. Leave it blank for a large population.

6. Click Calculate.

7. The calculator displays your Required Sample Size, along with the Z-score, margin of error, and population information.

When the calculated sample size is not a whole number, the required number of responses is rounded up to the next whole number.

How to Find Margin of Error

Use this option when you already know how many responses or observations you have.

1. Select Find Margin of Error.

2. Choose the Confidence Level.

3. Enter your Sample Size.

4. Enter the expected Population Proportion.

5. Enter the total Population Size when applicable, or leave it blank for a large population.

6. Click Calculate.

7. Review the calculated Margin of Error, Z-score, sample size, and population information.

How to Calculate Sample Size

For estimating a population proportion, sample size calculation is based mainly on the confidence level, desired margin of error, and expected population proportion.

For a large population, the basic sample size formula is:

$n_0 = \frac{Z^2 p(1-p)}{E^2}$

Where:

• (n_0) = required sample size

• (Z) = Z-score for the selected confidence level

• (p) = expected population proportion as a decimal

• (1-p) = remaining proportion

• (E) = margin of error as a decimal

For example, 50% is entered into the formula as (0.50), while a 5% margin of error is (0.05).

Sample Size Formula for a Finite Population

If the total population is known and finite, the initial sample size can be adjusted using the finite population correction:

$n=\frac{n_0}{1+\frac{n_0-1}{N}}$

Where:

• (n) = adjusted sample size

• (n_0) = sample size calculated for a large population

• (N) = total population size

This adjustment can reduce the required sample when the sample represents a meaningful portion of the entire population.

For a very large population, the adjustment becomes small. That is why the sample size calculator lets you leave Population Size blank when working with a large population.

Sample Size Calculation Example

Suppose you are conducting a survey and choose:

• Confidence level: 95%

• Margin of error: 5%

• Population proportion: 50%

• Population: Large

A 95% confidence level corresponds to a Z-score of approximately (1.96).

Using the formula:

$n_0=\frac{Z^2p(1-p)}{E^2}$

Substitute the values:

$n_0=\frac{(1.96)^2(0.50)(1-0.50)}{(0.05)^2}$

$n_0=\frac{3.8416\times0.25}{0.0025}$

$n_0=384.16$

Because you cannot collect a fraction of a response and rounding down could leave you below the required minimum, round the result up:

$\boxed{n=385}$

So, you need at least 385 responses under these assumptions.

How Is Margin of Error Calculated?

If you already have a sample size, you can estimate the margin of error for a population proportion.

For a large population, the formula is:

$E=Z\sqrt{\frac{p(1-p)}{n}}$

Where:

• (E) = margin of error

• (Z) = Z-score

• (p) = expected population proportion

• (n) = sample size

To express the result as a percentage:

$\text{Margin of Error (\%)} = E \times 100$

Margin of Error Example

Suppose:

• Confidence level = 95%

• Z-score = 1.96

• Sample size = 385

• Population proportion = 50%

• Population = Large

Then:

$E = 1.96\sqrt{\frac{0.50(1-0.50)}{385}}$

$E = 1.96\sqrt{\frac{0.25}{385}}$

$E \approx 0.04994$

Convert the decimal to a percentage:

$0.04994 \times 100 \approx 4.99\%$

So, the estimated margin of error is approximately 4.99%, matching the margin of error calculator result.

Common Confidence Levels and Z-Scores

A confidence level describes how often a confidence-interval procedure would capture the true population value over many repeated random samples under the model assumptions.

Common confidence levels and approximate Z-scores are:

A higher confidence level requires a larger Z-score. If all other inputs remain unchanged, this generally increases the required sample size.

For example, a study designed for 99% confidence will usually require more responses than the same study designed for 95% confidence.

Why Use a 50% Population Proportion When Unsure?

If you do not have a reliable estimate of the expected population proportion, 50% is commonly used as a conservative choice.

The sample size formula contains:

$p(1-p)$

When:

$p=0.50$

then:

$0.50(1-0.50)=0.25$

This is the maximum possible value of (p(1-p)). Therefore, using 50% produces the largest sample-size estimate for the same confidence level and margin of error.

For example:

$0.50(0.50)=0.25$

but:

$0.20(0.80)=0.16$

So, if you have no reliable prior information about the proportion, 50% provides a conservative starting point.

If trustworthy previous research suggests a realistic expected proportion, you may use that estimate instead.

How Confidence Level and Margin of Error Affect Sample Size

The number of responses you need can change significantly depending on the settings you choose.

Higher confidence level: Increasing the confidence level generally increases the required sample size because you are asking for greater confidence in the interval-estimation procedure.

Smaller margin of error: A narrower margin of error requires a larger sample. For example, targeting ±3% generally requires substantially more observations than targeting ±5% under otherwise identical assumptions.

Larger margin of error: Allowing less precision reduces the required sample size.

Population proportion: A proportion near 50% produces the greatest variability and therefore usually requires the largest sample under this formula.

Population size: Population size can matter when sampling without replacement from a relatively small, known population. For large populations, its effect on the required sample size becomes increasingly small.

Does Population Size Affect Sample Size?

Yes, but not always as much as people expect.

For a large population, the required sample size is driven mainly by your confidence level, margin of error, and expected population proportion rather than increasing directly in proportion to the total population.

For example, you do not necessarily need ten times as many responses simply because one population is ten times larger than another.

Population size becomes more important when you are sampling a substantial fraction of a known finite population. In that situation, the finite population correction reduces the required sample size.

This is why the margin of error calculator provides an optional Population Size field. Enter it when you have a specific finite population, or leave it blank when treating the population as large.

Common Sample Size Calculation Mistakes

Confusing Confidence Level With Margin of Error

These measure different things. Confidence level relates to the long-run coverage of the interval procedure, while margin of error describes its desired or estimated precision.

Rounding the Required Sample Down

If your calculation produces 384.16, using 384 would fall below the calculated requirement. Round upward to 385.

Automatically Using 50% When Better Information Exists

A 50% population proportion is useful when the expected proportion is unknown. If reliable prior research provides a suitable estimate, that value may give you a more study-specific calculation.

Ignoring the Finite Population

If you are sampling a substantial part of a small, known population without replacement, finite population correction can affect the required sample size.

Assuming a Larger Sample Fixes Every Research Problem

Increasing sample size can improve statistical precision, but it does not automatically eliminate problems such as selection bias, nonresponse bias, poor survey questions, measurement errors, or an unrepresentative sampling method.

Using This Formula for Every Type of Study

This margin of error calculator is designed around estimating a population proportion. Sample-size calculations for means, hypothesis tests, A/B tests, clinical trials, regression, clustered samples, or statistical power analysis may require different formulas and additional assumptions.

Tips for More Reliable Sample Size Planning

Start by clearly defining the population you want your findings to represent. A statistically adequate sample is much more useful when observations are selected using an appropriate sampling method.

Choose the margin of error based on the precision your research actually requires rather than simply selecting the smallest value possible. Smaller margins of error can dramatically increase the number of responses needed.

Use prior research to estimate the population proportion when trustworthy information is available. Otherwise, 50% is a practical conservative choice for this type of calculation.

Also remember that the required sample size means usable responses, not necessarily the number of people you need to contact.

For example, if you need 385 completed responses but expect only 50% of invited participants to respond, you may need to invite roughly:

$\frac{385}{0.50}=770$

This adjustment is about expected response rate and is separate from the statistical sample size formula.

Helpful Resources

• Confidence Interval Basics

Pro Tips

• The margin of error is the desired level of precision.

• The confidence level determines the level of confidence in the sample size.

• Higher confidence levels require larger sample sizes.