Simple Random Sample How To Calculate

Use this premium calculator to estimate the sample size needed for a simple random sample. It applies the standard proportion sample size formula and finite population correction so you can plan surveys, audits, quality checks, and research projects with confidence.

How to calculate a simple random sample

A simple random sample is one of the most important ideas in statistics because it gives every member of a population an equal chance of being selected. When researchers talk about survey accuracy, representativeness, and inference from a sample back to a population, simple random sampling is often the starting point. If you are trying to understand simple random sample how to calculate, there are really two practical questions you need to answer:

1. How large should the sample be?
2. How do you actually select units randomly once you know the sample size?

This page focuses first on sample size calculation, because that is usually what people mean when they ask how to calculate a simple random sample. The calculator above estimates the number of observations you need when measuring a proportion, such as customer satisfaction, support for a policy, defect rate, or prevalence of a trait.

What a simple random sample means

In a true simple random sample, each element in the population has the same probability of being selected, and each possible sample of a given size has the same chance of occurring. That may sound abstract, but in practice it means you need a complete list of the population and a random method to pick units from it. For example, if you have a list of 10,000 registered users, you can assign a number to each user and then use a random number generator to choose your sample.

The power of simple random sampling comes from fairness and mathematical tractability. Because selection is random, you can estimate sampling error and build confidence intervals. This is why federal agencies, universities, and major research organizations rely on probability sampling methods in formal studies.

The standard formula for sample size

For proportions, the standard starting formula is:

n0 = (Z² × p × (1 – p)) / e²

Where:

• n0 = the initial sample size estimate for a very large population
• Z = the z-score for your confidence level
• p = the estimated proportion expressed as a decimal
• e = the margin of error expressed as a decimal

If the population is finite, you then apply the finite population correction:

n = n0 / (1 + ((n0 – 1) / N))

Where N is the population size and n is the adjusted sample size. In practical use, you round up to the next whole number, because you cannot sample a fraction of a person or record.

What confidence level should you use?

The confidence level reflects how certain you want to be that your interval captures the true population value in repeated sampling. Common choices are 90%, 95%, and 99%. A 95% confidence level is the most widely used default in business, social science, public health, and opinion research. The associated z-scores are standard:

Confidence level	Z-score	Typical use case	Impact on sample size
90%	1.645	Exploratory internal studies, early market testing	Smaller sample than 95%
95%	1.960	Standard surveys, academic work, operational decisions	Common balance of precision and cost
99%	2.576	High-stakes risk analysis, compliance-sensitive projects	Largest sample of the three

Why 50% is often used for the estimated proportion

If you do not know the expected proportion, statisticians often use p = 0.50. This is a conservative choice because it produces the maximum variance for a proportion and therefore gives the largest required sample size. In plain language, using 50% helps protect you from underestimating how many observations you need.

For example, if you think the true proportion might be closer to 10% or 90%, the required sample size may actually be smaller. But if you are uncertain, 50% is the safest planning assumption.

Step by step example

Suppose you manage a customer database of 10,000 accounts and want to estimate the share of customers who would recommend your service. You want 95% confidence, a 5% margin of error, and you do not know the expected proportion, so you use 50%.

1. Convert the values to decimals where needed: p = 0.50, e = 0.05, Z = 1.96.
2. Compute the large-population estimate:
   n0 = (1.96² × 0.50 × 0.50) / 0.05² = 384.16
3. Apply finite population correction for N = 10,000:
   n = 384.16 / (1 + ((384.16 – 1) / 10000)) = about 369.98
4. Round up to a whole number: 370

That means a simple random sample of 370 customers is enough for the precision settings you chose. This is one reason people are often surprised that sample size does not grow in direct proportion to population size. Once the population becomes fairly large, the sample size needed for a given confidence level and margin of error changes much more slowly.

Comparison table: sample size by confidence and margin of error

The table below uses the standard large-population formula with p = 50%. These are widely recognized benchmark values in introductory statistics and survey methodology.

Confidence level	Margin of error	Large-population sample size	Rounded planning value
90%	5%	270.60	271
95%	5%	384.16	385
99%	5%	663.55	664
95%	4%	600.25	601
95%	3%	1067.11	1068

These numbers show a key statistical reality: narrowing the margin of error requires a much larger sample. Moving from 5% to 3% at 95% confidence does not just add a little more work. It almost triples the sample requirement. That is why project planning always involves a tradeoff between budget, timing, and precision.

Comparison table: effect of finite population correction

When your population is not extremely large, the finite population correction lowers the required sample size. Using the common benchmark of 95% confidence, 5% margin of error, and p = 50%, the unadjusted estimate is 384.16. After correction, the required sample changes as follows:

Population size	Unadjusted n0	Adjusted n	Rounded final sample
500	384.16	217.49	218
1,000	384.16	277.74	278
5,000	384.16	356.78	357
10,000	384.16	369.98	370
100,000	384.16	382.69	383

This table explains why finite population correction matters most for smaller populations. If your population is 500, the correction reduces the needed sample substantially. If your population is 100,000, the corrected value is almost the same as the large-population estimate.

How to draw the sample after you calculate the size

Calculating the sample size is only part of the job. To create a true simple random sample, you also need a random selection process. A practical workflow looks like this:

1. Create a complete sampling frame, meaning a list of every unit in the population.
2. Assign each unit a unique number from 1 to N.
3. Use a random number generator, spreadsheet random function, or statistical software to pick the required number of unique IDs.
4. Collect data from the selected units only, avoiding substitutions unless your methodology explicitly allows them.
5. Document nonresponse, exclusions, and replacements so the sampling process remains transparent.

If you substitute hand-picked cases for missing respondents, the sample is no longer a clean simple random sample. That is a common source of bias in real-world fieldwork.

Common mistakes people make

• Confusing population size with sample size. A larger population does not always require a dramatically larger sample.
• Using percentages instead of decimals inside the formula. For example, 5% must be entered as 0.05 in the formula.
• Ignoring the estimated proportion. If unknown, use 50% rather than guessing low and underpowering the study.
• Forgetting to round up. Always round upward for planning.
• Assuming random selection when convenience sampling was actually used. Statistical formulas only justify inference when sampling is genuinely random.
• Ignoring nonresponse. If you expect only 60% of selected people to respond, divide the required completed sample by 0.60 to estimate how many invitations to send.

When this calculator is appropriate

This calculator is best when your outcome is a proportion and your design is close to a simple random sample. It works well for:

• Customer surveys
• Employee opinion polls
• Audit sampling for pass or fail outcomes
• Quality assurance defect checks
• Student, patient, or resident proportion estimates

It is less appropriate when you are estimating a mean rather than a proportion, using complex designs such as clustering or stratification with unequal weights, or doing power analysis for experimental hypothesis tests. In those cases, specialized formulas are needed.

How nonresponse changes your field plan

Suppose the calculator tells you that you need 370 completed responses, but you expect only a 50% response rate. Then your contact list should include around 740 randomly selected units. This does not change the number of completed responses you need for precision, but it does change your operational plan. A lot of real studies fail because teams calculate the needed completed sample correctly and then forget to account for nonresponse.

Authoritative references for deeper study

If you want to validate your methodology or read more about probability sampling and survey accuracy, these authoritative sources are useful:

• U.S. Census Bureau: Simple Random Sampling
• Penn State University STAT 500: Applied Statistics
• CDC Principles of Epidemiology: Sampling and Survey Concepts

Final takeaway

When people ask simple random sample how to calculate, the answer is usually: define your population, choose a confidence level, choose a margin of error, select an estimated proportion, calculate the preliminary sample size, then apply finite population correction if needed. After that, use a truly random method to select the units. The formula is straightforward, but the quality of the final study depends just as much on disciplined execution as it does on correct arithmetic.

Use the calculator at the top of this page to estimate the sample size instantly. If you are unsure which settings to choose, start with 95% confidence, 5% margin of error, and 50% estimated proportion. That is the most common planning baseline and a strong practical starting point for many real-world surveys.