Variable Sampling
Variable sampling is a quality-control and acceptance-sampling method that uses measured numerical data from a sample lot or process, rather than pass/fail counts, to determine whether material meets specification. It applies to continuous characteristics such as length, thickness, weight, time, volume, and concentration, enabling decisions based on the sample's distribution rather than simple attribute counts.
On a manufacturing shop floor, variable sampling is most visible at receiving and incoming inspection. When a truckload of bar stock, coils, resin pellets, or packaged components arrives, a small random sample is pulled instead of checking every unit. The characteristic—diameter, thickness, moisture, gauge, or fill weight—is measured against specification limits, and the sample mean and standard deviation are compared with an acceptance constant. If the statistic satisfies the plan, the lot is released into inventory; if not, it is quarantined. The same logic extends to warehousing, where ANSI/ASQ Z1.9 plans permit checking statistical subsets of cartons or bundles for weight or dimensions. For ongoing process monitoring, variable sampling tracks the same continuous feature over time, detecting shifts in process mean or spread before defects reach finished goods. The result is faster, defensible disposition decisions, lower inspection costs, and tighter protection of downstream yield.
When is variable sampling preferred over attribute sampling?
Variable sampling is preferred when the quality characteristic is measurable on a continuous scale and the actual measurement can drive a lot-disposition decision, especially when sample sizes must be minimized because each observation carries more information than a pass/fail result.
What inputs define a variables acceptance sampling plan?
A variables acceptance plan is defined by sample size n, acceptance constant k, specification limits (LSL/USL), and whether process variability is known or estimated from the sample. The lot is accepted if the computed statistic meets the acceptance rule.
How does variable sampling support warehouse inventory disposition?
It converts a small set of measured samples into a statistically defensible release, hold, rework, or reject decision. Plans like ANSI/ASQ Z1.9 allow high-volume lots to be dispositioned quickly without full inspection, while still protecting downstream production from nonconforming material.