NP Chart (Number Defective)
Monitor count of defective items in processes with constant sample sizes for simplified interpretation.
Chart Overview & Key Characteristics
What is an NP Chart?
The NP chart (number defective chart) is an attribute control chart used to monitor the number of defective items in a process when the sample size remains constant. Unlike the P chart which tracks proportions, the NP chart directly plots the count of defective items, making it easier to interpret when sample sizes are fixed. It helps detect changes in the defect rate and overall process quality over time.
Key Characteristics
- • Purpose: Monitor number of defective items with constant sample size
- • Data Type: Attribute data (count of defectives)
- • Sample Size: Must be constant across all subgroups
- • Best For: Processes with fixed inspection quantities
Typical Applications
- • Manufacturing: Number of defective products per fixed batch size
- • Quality control: Defective items in samples of 100 units
- • Service: Number of errors in fixed-size document batches
- • Healthcare: Number of adverse events per fixed patient group
- • Software: Number of failed tests in fixed test suites
Control Limits & Statistical Foundation
Average defective proportion and average number defective across subgroups
npi
= number of defectives in subgroup i
= number of subgroups
= constant sample size
NP chart control limits
= average number defective
= overall proportion defective
Binomial distribution relationship
= sample size
= true proportion defective
Sample Size Requirements
- • Sample size must be constant for all subgroups
- • Minimum: np̄ ≥ 5 and n(1-p̄) ≥ 5
- • Recommended: n ≥ 50 for low defect rates
- • For very low p̄: may need n ≥ 1000
Reading the np Chart
Example Chart & Pattern Recognition
sixsigmacenter.com Implementation Guide
Platform Features for NP Charts
- • Simplified Setup: Easier than P charts due to constant sample sizes
- • Automated Validation: Ensures sample size consistency across subgroups
- • Constant Control Limits: No recalculation needed for each point
- • Direct Count Interpretation: Charts show actual defective counts for immediate understanding
Preparing Your Data for Upload
- • One row per inspection period, with a period label column and a count-of-defective-units column. Count units, not defects: a part with three flaws is one defective.
- • Keep the sample size genuinely constant across rows. If it varies, use a P chart instead — an NP chart with drifting n produces limits that mean nothing.
- • Record the sample size once in the setup rather than pre-computing proportions. The platform derives the center line and limits from the raw counts.
- • Keep the rows in production time order so run rules and shift patterns are detectable.
Size the subgroup so the average count defective is at least 5. Below that the normal approximation behind the three-sigma limits breaks down.
Platform Implementation on sixsigmacenter.com
Common Implementation Errors
- • Using variable sample sizes (should use P chart instead)
- • Not checking minimum sample size requirements
- • Confusing defectives with defects (use C chart for defects)
- • Forgetting to set lower control limit to zero when negative
Example Data Layout for NP Chart
Data for NP Chart with constant sample size. Count the number of defective items (not defects) in each fixed-size sample. Sample Size Recommendation: Fixed sample size of 50-100+ units with 20-25 subgroups for reliable control limits - constant sample size essential for np chart validity.
This is how your spreadsheet or CSV file should be structured before you upload it. Match the column names, keep one row per observation, and leave the analysis to the platform.
Process in Statistical Control
- • All counts fall within control limits
- • No systematic patterns or trends
- • Random variation around center line
- • Defect count is stable and predictable
Out-of-Control Indicators
- • Points beyond control limits (special causes)
- • 8 consecutive points above or below center line
- • Obvious trends or patterns in defective counts
- • Sudden shifts in process defect level
Action Guidelines
- • High counts: Investigate for root causes of defects
- • Increasing trend: Process deterioration, take corrective action
- • Decreasing trend: Process improvement, identify what changed
- • Stable but high np̄: Process needs fundamental improvement