UTS INSPECTIONYour quality control expert
  • Home
  • How Does Lot Size Affect the AQL Sample Size?

How Does Lot Size Affect the AQL Sample Size?

Lot size and inspection level determine a sample-size code letter. That letter, together with the agreed AQL and sampling scheme, leads to the sample size in the applicable table. As lot size increases, the sample generally grows in steps rather than as a fixed percentage of the lot; some increases leave the sample unchanged.

At UTS, we provide third-party product inspection, factory audit, supplier evaluation, and laboratory testing coordination. This guide explains how to define an inspection lot and select a sampling plan. Our reports present actual findings only, not corrective-action plans, improvement recommendations, or shipment-release advice.

What Counts as the Inspection Lot?

The lot is the defined group of units submitted for an acceptance decision. Its size is the number of units in that group, not automatically the purchase-order total or the factory's output. Identify the product, quantity, production conditions, and inspection location before selecting a sample.

Goods from different production conditions may need separate lots. Units added after selection were not represented in the original draw. Goods at another location require an agreed arrangement that makes the defined lot available for sampling; a result from accessible goods must not be extended silently to unavailable goods.

For our final random inspection, production is complete and at least 80% is packed. These are inspection-readiness conditions. They do not mean that only the packed 80% becomes the lot. Define the submitted lot separately and confirm access to the units it includes.

How to Find the AQL Sample Size

In the attribute-sampling systems discussed by NIST, the inspection level determines the relationship between lot size and sample size. There are three general levels and four special levels. The lookup begins with lot size and inspection level, then proceeds from the code letter to the applicable sampling-plan table. [1]

StepWhat to confirmResult
Define the lotProduct, submitted quantity, and availabilityThe lot-size range to use
Select the code letterLot-size range and agreed inspection levelThe starting code letter
Select the planStandard and edition, AQL, inspection severity, and single, double, or multiple samplingThe applicable plan and sample requirements
Read the table instructionsAny redirection to another plan and all associated notesThe sample size and decision rules actually used

The code letter is a starting point, not a substitute for reading the plan. Section 4.9.3 of the historical MIL-STD-105E explains that an unavailable AQL/code-letter combination can direct the user to another letter, whose sample size then applies. We cite this government-issued document to explain the mechanism, not as a claim that it is the current contractual standard. Use the edition agreed for your order. [2]

For a single-sampling plan written as (n, c), n is the number of units inspected and c is the acceptance number. The lot meets that plan's acceptance criterion when the observed count does not exceed c; a higher count fails that criterion. Keep the specified counting basis consistent: nonconforming units and individual nonconformities are not interchangeable. [3]

What Happens When the Lot Gets Larger?

The Sample Changes in Steps

Lot-size ranges work like stairs. At a fixed inspection level, an increase within the same range leaves the starting code letter unchanged. Moving into another range may change it. With the other plan inputs held constant, the resulting sample therefore need not increase each time the lot grows.

The Checked Percentage Can Fall

AQL sampling is not normally a fixed-percentage inspection. As a hypothetical arithmetic example, if the sample size stays unchanged while the lot grows from 1,000 to 10,000 units, the checked fraction becomes one-tenth as large. This illustrates n divided by lot size; it does not specify a valid sample size for either lot.

Table Limits Still Apply

The largest lot-size range in a table can cover further increases without another code-letter change. That does not make one sample size universal for all large orders. The level, AQL, severity, sampling type, and table instructions still matter. If the plan does not provide the discrimination the buyer needs, review the agreed sampling scheme before inspection rather than simply multiplying the sample by the increase in quantity.

What Lot Size Does Not Decide

  • The agreed AQL. For example, choosing an AQL of 2.5 for a defect class does not mean that any individual shipment is permitted to contain exactly 2.5% defective units. AQL is a plan-selection input, not the observed defect percentage.
  • The inspection level. General Level II is normally used unless another level is specified. Levels I and III provide different discrimination; special levels are intended for checks requiring relatively small samples, with the associated sampling risk considered.
  • Normal, tightened, or reduced inspection. Apply the scheme's switching rules and relevant inspection history. A larger lot alone is not an instruction to switch severity.
  • The acceptance number. Read Ac and Re from the applicable plan. Do not calculate them by multiplying the AQL percentage by n.

A larger sample alone does not establish a stricter acceptance decision. The sample size and acceptance number work together to determine the probability of accepting lots at different quality levels. Assess the plan as a whole rather than comparing n alone. [4]

Common Lot-Size Mistakes

AssumptionWhat to do instead
Doubling the lot requires twice the sampleLook up the new lot-size range with the same agreed plan inputs. The sample may stay unchanged or move to another size.
Splitting a lot on paper automatically reduces inspectionDefine genuine lots first. Separate lots require separate plan selection and acceptance decisions; their combined sample is not necessarily smaller.
Passing one sample also covers goods added laterKeep the result tied to the lot represented at selection. Newly added or unavailable goods are not automatically covered.

A Hypothetical Mug Order

Suppose a buyer increases an order for the same ceramic mug before inspection. First establish whether the additional goods belong to the same submitted lot. If they do, use the revised total quantity to look up the code letter again. If it is unchanged, and all other plan inputs remain unchanged, the sample requirement remains the same. If the letter changes, read the applicable plan again rather than scaling the old sample. This is a hypothetical decision example, not a UTS customer case or a table of sample-size values.

How We Connect Sampling to Inspection

Our published inspection workflow describes checks against the purchase order, specification, approved sample, drawings, and packing files. For sampled inspections, we use the buyer-approved inspection level and sampling plan. Our final random inspection service includes random sampling of finished goods and checks of quantity, workmanship, function, labels, and packing within the agreed scope.

For lot-size planning, those materials have distinct roles: the purchase order identifies the order, quantity information identifies the submitted goods, and the specification defines the checks. A random sample must come from the lot it is intended to represent. NIST describes acceptance sampling as using a random sample from a lot to decide its disposition. [5]

Our reports record actual inspection findings, such as quantity status, measurements, defects, test results, and checks that could not be completed. The preparation guidance here is separate from report content. A sampling result does not establish that every unit is defect-free or that all legal requirements have been met.

Our initial production inspection and during-production inspection address earlier production stages. Their scope should be agreed for that stage; they are not interchangeable with a finished-lot acceptance decision.

Where barcode or QR-code checks are included, specify the checked units, scanning device and conditions, and expected encoded information. Every checked code must scan successfully and match the approved data: the scan-success requirement is 100% for checked codes. This does not mean that the entire lot was scanned or that readability on every future device has been established.

Before You Confirm the Sample

Bottom line: a larger lot does not call for a fixed percentage to be inspected. Define the lot, find its code letter at the agreed inspection level, and read the complete sampling plan before confirming the sample size.

Need to define the sampling scope for your shipment?

Share your product specification, submitted lot quantity, packing status, and agreed sampling requirements with our team before the visit.

Contact Our Team

References

  1. [1] NIST/SEMATECH e-Handbook: Choosing a Sampling Plan, MIL Standard 105D.
  2. [2] U.S. Department of Defense: historical MIL-STD-105E, sections 4.9.1–4.9.3 and Table I. Document copy hosted by SQC Online.
  3. [3] NIST/SEMATECH e-Handbook: Lot Acceptance Sampling Plan Types.
  4. [4] NIST/SEMATECH e-Handbook: Choosing a Sampling Plan with a Given OC Curve.
  5. [5] NIST/SEMATECH e-Handbook: What Is Acceptance Sampling?

Tel

+852-61343425

Tel

+86 757-86783812

Tel

+86 571-87423201

Whatsapp