Attribute Control Charts

Comprehensive guide to control charts for categorical data including proportions, counts, and classifications.

Chart Overview

Attribute Statistics & Control Limit Formulas

Fraction/Count Defective (Binomial Family):

P chart control limits (variable sample size)

= average fraction defective across all subgroups

= subgroup sample size

NP chart control limits (constant sample size)

= average number defective across subgroups

= constant subgroup sample size

Defect Counts (Poisson Family):

C chart control limits (constant inspection unit)

= average count of defects per inspection unit

U chart control limits (variable inspection unit)

= average defect rate per unit

= size of the inspection unit for that subgroup

Key Assumptions

• • P/NP charts assume a binomial distribution with constant probability of defect
• • C/U charts assume a Poisson distribution with a constant rate of occurrence
• • Observations are independent across units and subgroups
• • LCL is set to 0 whenever the computed value is negative

Reading the Attribute Control Chart

Interpretation Guidelines

sixsigmacenter.com Implementation Guide

Chart Selection Decision Tree

Example Data Layout for Attribute Control Charts

Example data for P Chart (fraction defective). For NP Chart, use Sample_Size and Defective columns. For C/U charts, replace with defect counts and inspection units. Sample Size Recommendation: Sample size should be adequate for reliable statistical analysis - minimum 20-25 subgroups with np ≥ 5 for proper control limit calculation.

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.