Attribute Sampling
Meaning ~ Metric for assessing shipment conformity by counting binary non-conformities against fixed acceptance thresholds to decide lot acceptance or rejection.
Stratified Randomization
Meaning
Sampling selection technique divides a production lot into distinct sub-groups or strata before drawing random samples proportionally from each sub-group. A stratified randomization governs sample selection across multiple production shifts, sub-assembly lines, material batches or packaging pallets. It applies to complex manufacturing lots where quality variations may occur between different production runs or time periods.
The technique stops applying once sample selection completes and physical testing commences.
Sampling Strategy
Technicians divide batch inventory into logical sub-groups based on production time, pallet number or machine cavity. Random samples are drawn independently from each stratum to ensure representation across the entire production run. Stratified sampling prevents concentrated sampling from a single pallet, which masks variations across the batch.
Defect Detection
Dividing production runs into discrete sub-groups improves detection of localized process shifts or machine wear. Defect clusters restricted to a single pallet are isolated quickly during subgroup testing. Identifying specific sub-groups containing defects minimizes scrap volume by allowing targeted lot sorting.
Audit Validity
Quality audit protocols mandate subgroup sampling to prevent biased sample selection by factory operators. Sourcing contracts require suppliers to document sampling schemes on final inspection certificates. Valid random sampling across all production strata ensures defensible quality data during commercial disputes.