Acceptance Sampling Calculator
Evaluate lot acceptance using sample size, accept, and reject numbers. Determine pass/fail decisions for incoming and final inspections.
Perform a comprehensive measurement system analysis with %Study Variation, %Tolerance, and ndc calculations. Validate your measurement systems.
Good: Part Variation dominates (>90%). Poor: GRR dominates.
| Metric | Value | Acceptable | Marginal | Unacceptable | Status |
|---|---|---|---|---|---|
| %Study Variation | 22.5 | < 10% | 10-30% | > 30% | Marginal |
| %Tolerance | 92.7 | < 10% | 10-30% | > 30% | Unacceptable |
| NDC | 6 | >= 5 | 2-4 | < 2 | Adequate resolution |
| Parameter | AIAG Minimum | Recommended |
|---|---|---|
| Operators | 2 | 3 |
| Trials per operator | 2 | 3 |
| Parts | 5 | 10 |
| Confidence factor (k) | 5.15 | 5.15 or 6.00 |
| NDC threshold | 5 | 5+ |
Measurement System Analysis (MSA) is a structured approach to evaluating the statistical properties of a measurement system. While Gage R&R focuses on repeatability and reproducibility, a full MSA also considers bias, linearity, stability, and the number of distinct categories (ndc) the system can discriminate.
It gives a simplified MSA evaluation, computing %Study Variation, %Tolerance, and ndc from your measurement study data. These three metrics together give a comprehensive picture of whether your measurement system is adequate for its intended purpose โ process control, inspection, or both.
A capable measurement system is foundational to all quality decisions. Without it, process capability indices, SPC charts, and inspection results are all compromised by measurement noise.
MSA validates that your measurement data is trustworthy. It quantifies how much observed variation is real process variation versus measurement noise, enabling informed decisions about gage selection, calibration, and inspector training.
%Study Variation = (GRR / Total Variation) ร 100
%Tolerance = (GRR ร 5.15 / Tolerance) ร 100
ndc = 1.41 ร (Part Variation / GRR)
Acceptance: %SV or %Tol < 10% and ndc โฅ 5Result: %SV = 22.5%, %Tol = 9.3%, ndc = 6
The measurement system shows 22.5% study variation (marginal) but only 9.3% tolerance consumption (acceptable). ndc = 6 indicates adequate discrimination. Improvement would focus on reducing GRR to bring %SV below 10%.
1. **Bias** โ Difference between the observed average and the true reference value. 2. **Linearity** โ Consistency of bias across the operating range. 3. **Stability** โ Variation over time under constant conditions. 4. **Repeatability** โ Variation from the equipment under identical conditions. 5. **Reproducibility** โ Variation introduced by different operators or conditions.
The IATF 16949 automotive quality standard requires MSA for all measurement systems referenced in the control plan. This is audited during third-party assessments and applies to all measurement equipment used for product acceptance.
For pass/fail or visual inspection, attribute agreement analysis replaces the traditional Gage R&R. Operators evaluate the same parts and agreement percentages and kappa statistics are computed.
Last updated:
Gage R&R is one component of MSA focusing on repeatability and reproducibility. A complete MSA also evaluates bias (accuracy), linearity (accuracy across range), stability (consistency over time), and resolution.
Number of Distinct Categories (ndc) indicates how many groups the measurement system can reliably distinguish in the process data. ndc = 1 means the system only sees "one pile" โ no discrimination. ndc โฅ 5 is the minimum for adequate analysis.
Use %Study Variation when the measurement purpose is statistical process control. Use %Tolerance when the purpose is conformance inspection (pass/fail against spec). Many organizations evaluate both.
At minimum annually, and whenever a gage is recalibrated, repaired, or replaced. Also conduct MSA when operators change, environmental conditions change, or when process improvement requires higher measurement precision.
Yes. Minitab, JMP, and other quality software perform comprehensive MSA with ANOVA decomposition. It gives quick screening calculations for field use or when full software is not available.
Investigate root causes: gage condition, calibration, resolution, fixturing, operator technique, and environment. Address the largest variation contributor first. Retest after improvements to verify effectiveness.
Evaluate lot acceptance using sample size, accept, and reject numbers. Determine pass/fail decisions for incoming and final inspections.
Determine sample size and accept/reject numbers for incoming lot inspection based on AQL, lot size, and inspection level per ANSI/ASQ Z1.4.
Calculate AOQL โ the maximum average defect rate reaching customers after rectifying inspection. Evaluate worst-case outgoing quality levels.