GluePuddingTool

Sample Size Calculator

Compute the minimum sample size for statistical surveys using the Cochran formula, with finite population correction, confidence level comparisons, full calculation steps and practical advice.

FreeOnline Tool
Loading…

How to Use

Enter the population size N (use 0 for an infinite population), pick a confidence level (90%, 95% or 99%), a margin of error (1%, 3%, 5% or 10%) and an estimated proportion p (default 50% for the most conservative result). The minimum sample size is computed immediately. The results show the corrected sample size, the infinite-population size n0, the Z value, a comparison table across confidence levels, the full formula steps and a recommendation based on your inputs. Everything recalculates live, and 'Load sample data' fills in a typical configuration.

Features

  • Minimum sample size via the Cochran formula
  • Infinite population and finite population correction (FPC) modes
  • Confidence levels of 90%, 95% and 99% with multiple margins of error
  • Comparison table across confidence levels
  • Full calculation steps plus practical sampling advice

Use Cases

Customer satisfaction surveys
Decide how many questionnaires to send to a population of 10,000 to achieve 95% confidence and 5% margin of error
Quality inspection planning
Estimate the minimum number of units to sample from a production batch while balancing cost and statistical reliability
Market research sampling
Size your survey using an estimated proportion p, falling back to the conservative 50% when unknown
Teaching and review
Use the comparison table and calculation steps to explain how confidence levels affect sample size in a review meeting

FAQ

Which formula is used?
The Cochran formula: n0 = Z-squared times p(1-p) divided by e-squared. For a finite population, the result is adjusted with the finite population correction (FPC)
What should I enter for the proportion p?
If you have no estimate, keep the default 50%, which yields the largest, most conservative sample. Use a historical estimate when available
What does population size 0 mean?
It treats the population as infinite and skips the finite population correction, suitable when the population is far larger than the sample
Is the result a minimum?
Yes, it is the smallest sample that meets the chosen confidence and error requirements. In practice, add a buffer for invalid responses