p-Chart for Quality Control: When and How to Use Proportion Defective Charts

The p-chart (proportion defective chart) is one of the most practical quality control tools for manufacturers who work with attribute data — pass/fail inspections, go/no-go gages, or visual quality checks where parts are classified as defective or acceptable rather than measured on a continuous scale. If your quality data answers "how many parts out of this batch were bad?" rather than "what was the exact dimension of each part?", the p-chart is your tool.

For the broader SPC context and other chart types, see our guides on control charts and the quality control manufacturing guide.

When to Use a p-Chart

Use a p-chart when:

• Each item is classified as defective or non-defective (binary outcome)
• Sample sizes may vary between inspection periods (if sample size is constant, an np-chart is simpler)
• You are monitoring the overall defective rate of a process, not individual characteristics
• Sample sizes are large enough (50+ items) to produce meaningful proportions

Common Manufacturing Applications

• Final inspection pass/fail rates
• Visual inspection defective rates (cosmetic defects, surface flaws)
• Go/no-go gage results across a production batch
• Assembly verification (correct vs incorrect)
• Incoming material acceptance rates from suppliers
• Incoming quality inspection tracking

How to Build a p-Chart

Step 1: Collect Data

Record for each inspection period:

• n = number of items inspected
• d = number of defective items found
• p = d/n = proportion defective

Collect at least 20-25 data points (inspection periods) to establish reliable control limits.

Step 2: Calculate Average Proportion Defective

p-bar = total defective items / total items inspected = sum(d) / sum(n)

Step 3: Calculate Control Limits

For each sample of size n:

UCL = p-bar + 3 x sqrt(p-bar x (1 - p-bar) / n)

LCL = p-bar - 3 x sqrt(p-bar x (1 - p-bar) / n)

If LCL is negative, set it to zero (proportion defective cannot be negative).

Note: Because n can vary, control limits change for each sample. This creates the characteristic "stair-step" control limits of a p-chart.

Step 4: Plot and Interpret

Plot each sample's proportion defective (p) against the control limits. Look for:

• Points beyond control limits (special cause signal)
• Runs of points on one side of the center line
• Trends (six or more points consistently increasing or decreasing)
• Any non-random patterns per the Western Electric rules

Worked Example

A manufacturer inspects assembled units daily. Over 20 days:

p-bar = 68 / 2,150 = 0.0316

For Day 1 (n=100):

• UCL = 0.0316 + 3 x sqrt(0.0316 x 0.9684 / 100) = 0.0316 + 0.0525 = 0.0841
• LCL = 0.0316 - 0.0525 = 0 (set to zero since negative)

For Day 2 (n=120):

• UCL = 0.0316 + 3 x sqrt(0.0316 x 0.9684 / 120) = 0.0316 + 0.0479 = 0.0795
• LCL = 0 (same reasoning)

Simplifying Varying Control Limits

When sample sizes are similar (within 25% of each other), you can use the average sample size (n-bar) to calculate constant control limits. This simplifies the chart while remaining approximately correct.

n-bar = sum(n) / number of samples

This produces a single UCL and LCL pair, making the chart easier to plot and interpret. When a sample size deviates significantly from n-bar, calculate exact limits for that point.

Improving the Process: What to Do With p-Chart Data

A p-chart is not just for monitoring — it drives improvement:

When a Point Is Out of Control

1. Identify what was different about that production period (material lot, operator, machine, schedule disruption)
2. Conduct root cause analysis on the defective items
3. Implement corrective action through the CAPA process
4. Monitor subsequent data to verify the correction

When the Process Is in Control but Defective Rate Is Too High

The process is stable but not capable enough. Actions include:

• Use Pareto analysis to identify the top defect types
• Apply FMEA to prioritize improvement efforts
• Investigate whether scheduling issues contribute to defects (rushed setups, overtime, expediting)
• Implement process changes and recalculate control limits to verify improvement

When Defective Rate Decreases

Investigate what changed and standardize it. Recalculate control limits to reflect the new, improved process level. Tighter limits will detect future deterioration more quickly.

p-Chart vs Other Attribute Charts

Scheduling and p-Chart Performance

Track your p-chart data alongside production schedule events. Correlations are often revealing:

• Defective rates spike during weeks with heavy schedule disruptions
• Rush orders correlate with higher p-values
• Overtime shifts show higher defective proportions
• Properly scheduled production with adequate setup time produces lower, more stable defective rates

Use Spreadsheet QC to track p-chart data alongside scheduling metrics for integrated quality-scheduling analysis.

Frequently Asked Questions

Track Your Quality Data Effectively

Start monitoring your defective rates with p-charts using Spreadsheet QC. Pair quality tracking with RMDB scheduling to see how schedule stability impacts quality. Contact User Solutions for a complete quality and scheduling solution.