Calculate the minimum sample size needed for a survey or study, based on your desired confidence level and margin of error.
📖 Standard Sample Size Formula (with Finite Population Correction)Fill in the fields on the left with your information, then press Calculate to see your result instantly. No sign-up required, and no data is sent anywhere — everything is calculated right in your browser.
Before conducting any opinion poll or field study, it's essential to determine the minimum sample size needed to get statistically reliable, significant results. This calculator relies on the standard sample size formula, which takes into account the desired confidence level (90%, 95%, or 99%), the acceptable margin of error, and the estimated proportion of the studied characteristic in the population. If the total population size is known and finite, the calculator applies the "finite population correction" to reduce the required sample size compared to an assumed infinite population, because the sample represents a larger share of a relatively small population. Choosing a smaller margin of error or a higher confidence level always increases the required sample size, sometimes dramatically, which is exactly why real-world surveys must balance statistical precision against the practical cost and time required to reach a larger number of respondents. A small pilot study is often used beforehand specifically to obtain a more realistic proportion estimate, letting the final sample size calculation avoid the more conservative, larger 50%-assumption figure.
One of the most common practical questions in research design is deceptively simple to ask and genuinely important to answer correctly: how many people actually need to be surveyed for the results to be statistically trustworthy? Guessing too low produces unreliable results; guessing too high wastes time and money collecting more data than the study actually needs.
The standard sample size formula balances three inputs against each other. The confidence level (typically 90%, 95%, or 99%) reflects how certain the researcher wants to be that the sample results reflect the true population value within the stated margin of error — a higher confidence level requires a larger sample, all else equal. The margin of error is the acceptable range of uncertainty around the estimate — a survey aiming for a tighter margin of error (say, ±2% instead of ±5%) requires a substantially larger sample, since the relationship is not linear: cutting the margin of error in half roughly quadruples the required sample size.
The estimated population proportion matters because the formula is most demanding — requiring the largest sample — when the true proportion is close to 50%, since that is the point of maximum statistical variability. If prior research or a pilot study suggests the true proportion is closer to 10% or 90%, a smaller sample can achieve the same confidence and margin of error, which is why researchers often use the more conservative 50% assumption only when no better estimate is available.
The finite population correction addresses a specific scenario: when sampling from a small, known total population — surveying every employee at a 400-person company, for instance — a smaller sample is actually needed than the standard formula would suggest, because the sample already represents a substantial fraction of the entire population rather than a tiny sliver of an effectively infinite one. This correction becomes negligible once the population is very large relative to the sample, which is why it is typically ignored for national-level surveys but matters for smaller, well-defined populations.
Getting sample size right before data collection begins — rather than gathering a convenient but statistically inadequate number of responses and hoping for the best — is standard practice in market research, clinical trials, and academic social science, precisely because an underpowered study risks reaching confident-sounding conclusions that do not actually hold up under proper statistical scrutiny.
50% produces the most conservative (largest) required sample size when the true proportion is unknown, ensuring the study remains adequately powered.
It reduces the required sample size when your total population is small and known, since a sample represents a larger share of a small population.
Common choices are 5% for general surveys or 1–3% for studies requiring higher precision; smaller margins require larger sample sizes.