Overview
ISO 22514-5:2019 - part of the ISO 22514 series on statistical methods in process management - defines methods to estimate process capability and performance for attributive (attribute) characteristics. The standard describes how to calculate capability/performance indices when quality is expressed as counts or proportions (presence/absence, defect counts), complementing capability calculations for variable characteristics.
Key topics
- Attribute data models: Assumes binomial (proportion/nonconforming units) or Poisson (count/nonconformities) models and requires monitoring a single parameter for stability.
- Attribute control charts: Use of np, p, c and u charts to plot subgroup proportions or counts and to check statistical control before capability assessment.
- Process quality level (Q): Defines Q as the centre-line (average level) of a stable attribute control chart (p, np, c or u), and gives procedures for estimating Q from samples or lot data.
- Performance vs capability: Distinguishes process capability (requires statistical control) from process performance (no control-chart requirement). If the process is not in control, indices represent performance (P) rather than capability (C).
- Indices and measures: Introduces indices for attributive characteristics (Pp, Ppk, etc.) and related metrics such as first run capability (FRC), nonconformities per hundred units (NHU) and nonconformities per million (NMU / ppm).
- Confidence intervals and small-sample rules: Procedures for handling samples with no nonconformities (use of one-sided confidence limits) and for estimating upper confidence limits on Q when d = 0 or small d. Example: n = 200 with no defects → upper confidence limit Q ≈ 1.5% (illustrative in the standard).
- Preconditions: Guidance on sample size, subgrouping, specification definitions and treating production lines or locations separately.
Applications
- Estimating capability and performance when characteristics cannot be measured continuously (e.g., pass/fail checks, defect counts, visual inspections).
- Supplier evaluation and incoming quality assessment where attribute data are predominant.
- First-run capability estimates and production release decisions using NHU or ppm measures.
- Quality system monitoring in manufacturing, healthcare, software QA, inspection processes and audits that rely on attribute sampling.
Who should use it
- Quality engineers, process improvement specialists, statisticians, auditors, supplier quality managers and regulators who need standardized, statistically defensible methods to quantify capability/performance from attribute data.
Related standards
- ISO 22514 series (other parts cover variable characteristics and related methods)
- ISO 3534-2 (statistics - vocabulary and symbols)
- ISO 7870 series (control charts) and ISO 22514-7 (measurement uncertainty) are referenced for charting and specification aspects.
Keywords: ISO 22514-5, process capability, attributive characteristics, attribute control chart, process performance, NHU, FRC, p-chart, np-chart, c-chart, u-chart, nonconforming units.
Open preview (PDF)Statistical methods in process management — Capability and performance — Part 5: Process capability estimates and performance for attributive characteristics