This page is for anyone who has to look at a real fault on a real garment and decide which of three tallies it belongs in, because an inspection does not compare one defect count against one acceptance number; it keeps three separate tallies, one per defect class, each judged against its own acceptance number from the sampling table.
This matters most to a quality inspector, since a fault correctly counted but wrongly classified can pass an inspection that should have failed, or fail one that should have passed. Cedar & Co., a domestic brand selling to its own 12 shops from its own forecast, is used throughout this lesson: its own quality lead runs a final inspection on a stock batch of T-shirts supplied by Delta Garments before the batch leaves for the shops.
Without a clear, consistent classification, two inspectors counting the same faults on the same batch can reach two different pass/fail results, not because they disagree on what they saw, but because they disagree on what class it belongs in.
You need: lesson 9.2.2's own explanation of acceptance and rejection numbers, one per defect class.
By the end of this lesson you can:
Not covered here: what a factory, a buyer, or a buyer's agent does once a tally has actually crossed a rejection number (lesson 9.2.4), and how all three tallies are assembled into one signed report (lesson 9.2.5).
| Term | Meaning |
|---|---|
| Critical defect | A defect that could harm the user or breaks a legal or safety requirement; standard practice sets its acceptance number at zero. |
| Major defect | A defect likely to reduce the product's usability or its ability to be sold at full price, though it does not create a safety risk. |
| Minor defect | A defect unlikely to reduce the product's usability, appearance or sale value beyond a small tolerance most customers would not notice. |
| Defect count | The running total of defects found within one class, kept separately for critical, major and minor throughout one inspection. |
| Rejection number | The smallest number of defects of one class a sample can contain that fails the whole lot. |
An inspection does not ask "how many faults did we find" as a single question. It asks the question three times, once per class: how many critical defects, how many major defects, how many minor defects. Each of these three counts is compared only against the acceptance number the sampling table set for that class, at the AQL level agreed for it. A lot fails the moment any one of the three counts exceeds its own class's acceptance number, regardless of how the other two classes stand.
A person who has only classified faults on one product should not assume the same fault always lands in the same class everywhere. What differs is the standard applied, not the three-class structure itself.
| What can differ | The shapes you meet | Why it matters |
|---|---|---|
| The same fault type on different products | A loose thread on a basic T-shirt is typically a minor defect; the same loose thread near a small child's reach on a children's garment can be classed major or even critical if it risks detaching as a choking hazard. | The fault itself does not fix its class; the product and the risk it creates do. |
| A buyer's own written classification guide | An export buyer such as Northline Apparel often supplies its own defect classification guide in the buyer manual, listing exactly which faults it treats as major versus minor for its own garment types. | An inspector following a generic classification instead of the buyer's own guide can misclassify a fault the buyer would have judged differently. |
| A domestic seller setting its own standard | A domestic brand such as Cedar & Co., with no outside buyer's guide to follow, sets its own classification standard for its own stock, based on what its own customers would notice and tolerate. | The three-class structure and the logic of comparing each tally to its own acceptance number apply the same way, even with no buyer contract behind it. |
| A fault whose severity depends on its exact location | A small shade variation in a hidden inner facing is often minor; the same shade variation across a visible front panel, seen at first glance in a shop, is more likely major. | The same physical fault, measured the same way, can land in a different class depending on where on the garment it appears. |
Cedar & Co. buys finished T-shirts from Delta Garments on a 60-day open account, for its own shops, to its own forecast rather than one buyer's confirmed order. Before an illustrative batch of 4,000 T-shirts leaves for the shops, Cedar & Co.'s own quality lead runs a final inspection under its own AQL standard, since no outside buyer's contract sets one for a Make to Stock domestic batch like this.
| Fact | What it tells the quality lead |
|---|---|
| The batch is 4,000 pieces, an illustrative domestic stock quantity | The batch's own size, not Order A's, is what the sampling table would be entered with for this inspection |
| Cedar & Co. sets its own defect classification standard, with no outside buyer's guide | The quality lead applies Cedar & Co.'s own judgement of what its shop customers would notice and tolerate |
| Three separate tallies are kept during the inspection: critical, major, minor | A fault correctly identified but placed in the wrong tally can flip a pass into a fail, or the reverse |
| The batch is for Cedar & Co.'s own shops, not a single external buyer | The classification standard still needs to be written down and applied consistently, even without a buyer's audit to satisfy |
Cedar & Co. inspecting its own stock, with no external buyer's contract behind the inspection, is exactly why its own written classification standard, applied consistently, matters as much here as a buyer's own guide would on an export order.
The worked numberIn this illustrative inspection of the 4,000-piece batch, the quality lead finds three faults during the same pass through the sample: a loose sleeve hem thread (classed minor, since it does not affect wear and most customers would not notice it), a skipped stitch on a side seam long enough to open with wear (classed major, since it reduces the garment's usability before the first wash), and a sharp needle fragment left inside a collar seam (classed critical, since it could injure the wearer). Three faults, three different classes, three different tallies, each compared against its own acceptance number rather than a single combined count of three.
A common mistake is an inspector, seeing a major tally close to its acceptance number, quietly reclassifying a borderline fault as minor instead, so the batch still passes. This does not change what the fault actually is or what it will do once the garment is worn; it only changes which tally records it, which means the batch's real quality is not what the report now says it is. A batch passed this way can return as a customer complaint or a return, at a cost the saved inspection time never covered.
Every fault found in an inspection belongs to one of three groups: critical, if it could harm someone or breaks a safety rule; major, if it makes the product harder to use or sell at full price; or minor, if it is small enough that most customers would not notice or mind. The inspection keeps a separate running count for each group and compares each count only to its own limit. One serious fault in the critical group can fail the whole batch even if the other two groups look fine.
Cedar & Co.'s quality lead is finishing the final inspection of the 4,000-piece batch. The major tally already stands close to its acceptance number, and the shop manager, keen to get stock on the shelves for the weekend, suggests recording the newly found skipped-stitch fault as minor instead of major, since "it's basically the same either way."
The shop manager's note reads:
Should Cedar & Co.'s quality lead record the skipped-stitch fault as minor to keep the major tally under its acceptance number.
No. The fault must be classed according to Cedar & Co.'s own written standard, which defines it as major, regardless of which classification keeps a tally under its limit.
First, reclassifying the fault does not change what it actually is. A skipped stitch long enough to open further with wear will still open further with wear, whichever tally records it; the garment's real quality is unaffected by which column the fault is written into.
Second, the batch's own major tally, correctly recorded, sits exactly at its acceptance number. This means the batch is right at its own limit for major defects, a fact the quality lead needs to see clearly, not hide by moving one fault into a different tally.
Cedar & Co.'s quality lead should record the fault as major, report the batch's true position against its own acceptance number, and let that honest count decide whether the batch passes, rather than adjusting the classification to force a pass.
If Cedar & Co. ships the 4,000-piece batch to its shops with the skipped-stitch fault recorded as minor instead of major, and other garments with the same undetected fault type reach customers, the cost shows up later as returns and complaints rather than at the shop's own receiving dock, where it would have been cheaper to catch.
A return handled after a customer complaint costs the original making cost, the shop's handling time, and a share of the customer's trust in the brand, all of which are larger than the cost of accepting an honest major tally and deciding on the batch accordingly at the point of final inspection.
Open the AQL final inspection report builder's Defect Tally sheet. Enter each found fault's class, and watch the sheet's own formula compare that class's running count against its own acceptance and rejection numbers. This lesson's job is only to show why the class entered for each fault matters as much as the fault itself; lesson 9.2.5 shows how this sheet's own result feeds the full inspection report.
1. How many separate defect tallies does an AQL inspection keep, and why?
2. A batch has zero critical and zero major defects, but its minor tally exceeds its own acceptance number. What happens to the batch?
3. Why can the same fault type, such as a loose thread, sit in a different class on two different products?
| EN | AR | FR | TR | Meaning |
|---|---|---|---|---|
| Critical defect | عيب حرج | Défaut critique | Kritik kusur | A defect that could harm the user or breaks a legal or safety requirement; standard practice sets its acceptance number at zero. |
| Major defect | عيب رئيسي | Défaut majeur | Majör kusur | A defect likely to reduce the product's usability or its ability to be sold at full price, though it does not create a safety risk. |
| Minor defect | عيب ثانوي | Défaut mineur | Minör kusur | A defect unlikely to reduce the product's usability, appearance or sale value beyond a small tolerance most customers would not notice. |
| Defect count | عدد العيوب المرصود | Décompte des défauts | Hata sayımı | The running total of defects found within one class, kept separately for critical, major and minor throughout one inspection. |
| 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 classify and count a defect toward a pass or fail decision. The next lesson opens what happens once a lot's count crosses the rejection number: the options a factory, a buyer, or a buyer's agent has once a final inspection has actually failed. Back to the course map