Process Capability Analysis

Cp and Cpk use within-subgroup standard deviation (estimated from R̄/d₂ or S̄/c₄) — they represent inherent process capability assuming stability. Pp and Ppk use overall standard deviation (all data pooled) — they represent actual process performance including any instability. Both are displayed side-by-side with the gap flagged when significant.

Default color-coding: Red (< 1.00), Yellow (1.00-1.32), Green (1.33-1.66), Dark Green (≥ 1.67). Thresholds are configurable per characteristic to match customer-specific requirements. Automotive OEMs often require ≥ 1.67 for critical dimensions; Six Sigma targets ≥ 2.00.

Every Cpk and Ppk value includes a 95% confidence interval. When sample size is small (< 30 measurements), the system flags: 'Cpk confidence interval is wide — collect at least 30 measurements for a reliable estimate.' This prevents premature capability claims based on insufficient data.

An automatic Anderson-Darling normality test runs on every capability calculation. When data fails the normality check, SigmaResolve warns that standard Cpk may overstate or understate true capability and suggests non-normal analysis (Box-Cox or Johnson transformation) for accurate results.