Calculate the sample size needed to detect or confirm defect rates for pass/fail data — with zero-acceptance sampling, proportion estimation, and observed-defects confidence interval analysis.
Choose the type of attribute analysis you need to run.
Choose your analysis type, enter the parameters, and press Calculate to see the sample size, confidence intervals, and actionable guidance.
Launching a new component with a 1% defect rate assumption. Enter the lab and find how many items the C=0 plan requires you to inspect.
Use the calculator above to size pass/fail (attribute) samples for defect-rate confirmation and acceptance sampling. Includes zero-acceptance plans, proportion-estimation sizing and confidence intervals on observed defect rates — the standard toolkit for quality audits, supplier inspection and incoming-goods checks.
Attribute sample size answers the question "how many pass/fail items must I inspect to be sufficiently sure of the underlying defect rate?". It is the discrete counterpart to continuous-data sample-size calculations and is the standard approach for audits, supplier acceptance and conformity inspection.
For zero-acceptance: n = ln(1 − Confidence) ÷ ln(1 − AQL). For proportion estimation: n = (Z² × p(1−p)) ÷ E². For confidence on observed defects, the calculator uses exact binomial methods rather than the normal approximation, which matters when defect counts are small.
A buyer wants 95% confidence the supplier’s defect rate is below 1%. Zero-acceptance sample size: n = ln(0.05) ÷ ln(0.99) ≈ 299 units. Inspect 299 with zero defects allowed; any defect rejects the lot.
If the buyer relaxes the AQL to 2%, sample size drops to 149. Doubling the AQL roughly halves the sample. The calculator makes the trade-off between acceptable quality level and inspection cost explicit before the audit is commissioned.
Right-sized attribute sampling protects you against accepting bad batches and against over-inspecting good ones. It saves money on labour and reduces the defect-rate noise in supplier scorecards.
Use this for incoming-goods inspection, audit sample selection, supplier qualification, process audits, and any pass/fail acceptance decision.
Attribute sample size pairs with DPMO and Sigma Level to give a rigorous defect-rate measurement framework. Most ISO and IATF standards specify attribute-sampling requirements.
Define your AQL and confidence requirement before sampling. Use zero-acceptance plans for critical attributes; use proportion-estimation sizing when you genuinely need to estimate the rate.
Combine attribute sampling with control charts (p-chart, np-chart) and DPMO for a complete defect-rate measurement framework.
Sampling pass/fail (discrete) data, as opposed to continuous measurements. Each item is classified as conforming or defective, and the sample size depends on the defect rate you want to detect or confirm.
A plan in which any defect rejects the entire lot. It produces small sample sizes for a given confidence level and is the standard for critical-to-quality attributes.
Acceptable Quality Level — the worst defect rate that should still routinely be accepted. It is a planning input, not a target.
Continuous-data sampling uses standard deviation; attribute sampling uses defect rate (proportion). The formulas are different and the assumptions are too.
No. 30 is a heuristic for continuous data, not attribute data. Attribute samples for low defect rates often need hundreds to give meaningful confidence.
Want to understand attribute sampling in the context of a full measurement plan? The Green Belt covers this in full.
View Green Belt →