This page is for anyone who has agreed to run an AQL inspection and now needs the two numbers the method actually runs on: how many pieces to pull, and how many defects the pull may contain before it fails. Neither number is a judgement call. Both come from a published sampling table, entered with the lot's own size and the inspection level the buyer or the seller has set.
This matters most to a quality inspector or a merchandiser at a factory, since an inspection run with the wrong sample size, or the wrong acceptance number, is not a valid AQL inspection at all, whatever else it did correctly. Delta Garments' final inspection of Order A, its 15,000-piece T-shirt order, is used throughout this lesson to show exactly how the table is entered and read.
Without reading the table correctly, an inspector can pull a sample too small to be statistically meaningful, or apply an acceptance number meant for a different lot size, and report a result that looks official but was never actually the inspection the buyer agreed to.
You need: lesson 9.2.1's own explanation of why a factory samples instead of inspecting every piece.
By the end of this lesson you can:
Not covered here: how a found defect is classified as critical, major or minor before its count is compared against the acceptance number (lesson 9.2.3), and what happens once a lot's count exceeds the rejection number (lesson 9.2.4).
| Term | Meaning |
|---|---|
| Inspection sample size | The exact number of pieces an AQL inspection pulls from a lot, read from the sampling table using the lot's own size and the inspection level, not chosen by the inspector. |
| Sample size code letter | The letter the sampling table assigns to a lot size range and an inspection level, used to look up the exact sample size in the same table. |
| General inspection level | The standard inspection strictness (usually level II) a sampling table applies unless the buyer's contract states a different level. |
| Acceptance number | The largest number of defects of one class a sample may contain while the whole lot still passes. |
| Rejection number | The smallest number of defects of one class a sample can contain that fails the whole lot. |
A sampling table is read in a fixed order, never guessed at. First, the lot's own size is matched to one lot-size range in the table, which points to a sample size code letter. Second, that code letter is matched, at the agreed inspection level, to an exact sample size. Third, for each defect class being judged (critical, major, minor), the sample size row gives an acceptance number and a rejection number at the AQL level set for that class. Every one of these values comes from the same published table; none of them is set by the inspector's own judgement.
A person who has only read one row of the table should not assume every inspection lands on the same row. What differs is the row entered, not the three-step method used to enter it.
| What can differ | The shapes you meet | Why it matters |
|---|---|---|
| A larger or smaller lot than Order A's 15,000 pieces | A smaller export order lands in a lower lot-size range, giving a smaller sample size code letter and a smaller sample; a much larger order lands in a higher range, needing a larger sample. | An inspector who reuses Order A's own sample size on a different-sized lot is not running a valid inspection for that lot. |
| General versus special inspection levels | Most apparel orders use general inspection level II; a buyer requiring a stricter check for a sensitive product can specify a special level, which reads a different sample size from the same lot size. | The inspector cannot assume level II applies without checking the buyer's own contract or manual. |
| A tighter AQL for a safety-related defect class | A children's-wear order, or an order with metal trims near a child's reach, can set a stricter AQL for its critical defect class than a typical adult garment order, without changing the major or minor levels. | The same sample can be judged against different acceptance numbers for different defect classes at the same time. |
| Domestic sellers setting their own inspection level | A domestic brand such as Cedar & Co., with no outside buyer's contract to follow, can choose to apply general level II to its own finished stock as a matter of its own quality policy, reading the table exactly the same way an export order would. | The table itself does not care who is inspecting; only the inputs (lot size, inspection level, AQL per class) change who ends up with which row. |
Order A's lot size is 15,000 pieces. Northline Apparel's purchase order specifies general inspection level II, the level used for most apparel orders unless stated otherwise, and sets AQL 0 for critical defects, AQL 2.5 for major defects, and AQL 4.0 for minor defects, an industry-standard combination for a basic garment order. Delta Garments' inspector enters these figures into the sampling table, not into a private spreadsheet of remembered numbers, to find Order A's own row.
| Fact | What it tells the inspector |
|---|---|
| Order A's lot size is 15,000 pieces | This lot size falls in the table's 10,001-to-35,000-piece range |
| That range, at general inspection level II, carries sample size code letter M | Code letter M is the row the inspector reads the sample size from |
| Northline Apparel's purchase order sets AQL 0 / 2.5 / 4.0 for critical / major / minor | Three different acceptance and rejection numbers apply to the same one sample, one per defect class |
| The sample size and every acceptance/rejection number are fixed before counting starts | The inspector cannot adjust any of these numbers once defects begin turning up |
Order A being a single, confirmed export lot, inspected once against Northline Apparel's own stated levels, is exactly why its row in the sampling table is fixed and repeatable, not re-decided at each inspection.
The worked numberFollowing the widely published ANSI/ASQ Z1.4 single sampling table (general inspection level II, normal inspection), a lot size of 15,000 pieces carries sample size code letter M, which reads to a sample size of about 315 pieces. At that sample size, AQL 0 for critical defects gives an acceptance number of 0 and a rejection number of 1 (a single critical defect always rejects the lot, at any sample size); AQL 2.5 for major defects gives an acceptance number of about 21 and a rejection number of about 22; AQL 4.0 for minor defects gives an acceptance number of about 34 and a rejection number of about 35. These figures follow the standard, publicly documented table and should be checked against a current official copy before use in a real inspection.
A common mistake is an inspector reusing a sample size and acceptance number remembered from a previous, similarly sized order, without re-checking the current lot's exact size against the table. A lot size close to a range's boundary, such as 10,050 pieces against a boundary at 10,001, still falls in the same range as Order A's 15,000, but a lot at 9,900 pieces falls in the row below, with a different code letter and a different sample size. Reusing the wrong row means the sample size and every acceptance number are wrong, even if the inspector's counting of defects afterward is careful and accurate.
An AQL sampling table turns a lot's own size into three things: how many pieces to check, and how many faults are allowed for the class of fault being judged, one number for the largest allowed and one for the smallest that fails the lot. You find these by looking up the lot size, then the inspection level, then the required strictness for each type of fault. None of these numbers is a guess or a memory from a different order; each one is read from the table itself, freshly, for the lot in front of you.
Mona, the merchandiser at Delta Garments, is briefing the inspector before Order A's final inspection. The inspector, who ran a final inspection on a similarly sized order last month, suggests reusing that order's sample size of 280 pieces to save time, since "it was close enough last time."
The inspector's note reads:
Should Mona let the inspector reuse last month's sample size of 280 pieces for Order A's own 15,000-piece lot.
No. Order A's own lot size must be read fresh from the sampling table, because 15,000 pieces and 9,800 pieces fall in different rows with different sample sizes.
First, 9,800 pieces and 15,000 pieces sit on opposite sides of the table's row boundary at 10,001 pieces. A lot size close to a boundary, or simply in a different range, reads a different sample size code letter, even when the two orders feel similar in scale.
Second, reusing 280 pieces on Order A's own 315-piece requirement means the sample is undersized for this lot. An undersized sample is not a smaller, faster version of a valid AQL inspection; it is not a valid AQL inspection at all, whatever the count inside it shows.
Mona should have the inspector read Order A's own lot size fresh against the table before the inspection starts, confirming the sample size code letter and the sample size before a single piece is pulled.
If Delta Garments inspects Order A with a 280-piece sample instead of the 315 pieces its own lot size requires, and Northline Apparel later checks the inspection record against the sampling table, the inspection can be rejected as invalid regardless of the result it produced, because the sample itself did not meet the agreed method.
An invalid inspection record found after shipment forces Delta Garments to prove the lot's quality some other way, often a fresh inspection under time pressure at the buyer's own cost and schedule, which costs far more than the short time saved by skipping a fresh look at the table.
Open the AQL final inspection report builder's Sampling Plan sheet. Find the row for Order A's own lot size range, and read its sample size code letter, sample size, and the acceptance/rejection numbers for each defect class. This lesson's job is only to show how that one row is found and read; lesson 9.2.5 shows how these figures carry through into the full inspection report.
1. What three things does reading a sampling table correctly give an inspector?
2. A lot of 9,900 pieces and a lot of 15,000 pieces use the same inspection level. Why might they use different sample sizes?
3. Why does a critical defect class set at AQL 0 reject a lot the moment one critical defect is found, at any sample size?
| EN | AR | FR | TR | Meaning |
|---|---|---|---|---|
| Inspection sample size | حجم عينة الفحص | Taille d'échantillon d'inspection | Kontrol numune büyüklüğü | The exact number of pieces an AQL inspection pulls from a lot, read from the sampling table using the lot's own size and the inspection level, not chosen by the inspector. |
| Sample size code letter | حرف رمز حجم العينة | Lettre code de taille d'échantillon | Numune büyüklüğü kod harfi | The letter the sampling table assigns to a lot size range and an inspection level, used to look up the exact sample size in the same table. |
| General inspection level | مستوى الفحص العام | Niveau de contrôle général | Genel kontrol seviyesi | The standard inspection strictness (usually level II) a sampling table applies unless the buyer's contract states a different level. |
| Acceptance number | رقم القبول | Nombre d'acceptation | Kabul sayısı | The largest number of defects of one class a sample may contain while the whole lot still passes. |
| Rejection number | رقم الرفض | Nombre de refus | Ret sayısı | The smallest number of defects of one class a sample can contain that fails the whole lot. |
You now know how to read a sampling table to find a sample size and acceptance number. The next lesson opens the other half of the decision: how a found defect is classified as critical, major or minor, and why that classification changes which acceptance number applies. Back to the course map