Attribute control charts

Learning objectives

• Explain the purpose of attribute control charts in a garment-factory improvement project.
• Apply the described method to a representative shop-floor situation.
• Recognize the common mistakes and how to avoid them.

Concept

Attribute charts monitor proportions or counts when the response is pass/fail or number of defects.

p = defective proportion with variable n; np = defective count with constant n; c = defect count with constant opportunity; u = defects per unit with variable opportunity.

One garment may contain several defects, so defective and defect charts answer different questions.

Variables, units, and assumptions

Numerator = defectives or defects; denominator = sample size n or opportunity area.

Binomial (p/np) or Poisson (c/u) reasonably holds; independent items.

Free-spreadsheet workflow

Free spreadsheet: p = defectives/n per subgroup; limits = pbar ± 3·sqrt(pbar(1-pbar)/n_i).

Interpretation and limitations

Points outside limits or runs indicate special cause tied to definition.

Rare-event data may need geometric or g/h charts; definition drift invalidates baseline.

Garment-factory example

Use a p chart for end-line rejected garments when hourly sample size changes; a u chart for defects per inspected garment.

Method

1. Validate operational definitions.
2. Record numerator and denominator/opportunity.
3. Calculate center and sample-specific limits.
4. Plot.
5. Investigate signals with cause evidence.

Common mistakes

• Charting defect count without an exposure denominator.
• Switching defect definitions mid-baseline.