Study guide

8.3 Statistical Process Control & Process Capability

ASQ Certified Manager of Quality/Organizational Excellence CMQ/OE

Track your progress as you read

Finished sections get a check mark.

1 Chapter 0: Introduction & Exam Overview

0.1 Current CMQ/OE Exam Facts 0.2 Study Strategy by BoK Weight

2 Chapter 1: Leadership Principles & Organizational Structures

1.1 Leadership Theories & Styles 1.2 Organizational Structures & Matrix Management 1.3 Change Management Models & Strategies 1.4 Organizational Culture & Quality Governance

3 Chapter 2: Teams, Facilitation & Ethics

2.1 Team Formation & Development Stages 2.2 Facilitation Techniques & Team Empowerment 2.3 Conflict Resolution & Negotiation Strategies 2.4 ASQ Code of Ethics & Legal Compliance

4 Chapter 3: Strategic Planning & Environmental Analysis

3.1 Strategic Planning Models & Environmental Scans 3.2 Strategic Risk Management & Scenario Planning 3.3 Portfolio Management & Strategic Project Alignment

5 Chapter 4: Policy Deployment & Performance Alignment

4.1 Hoshin Kanri & Catchball Process 4.2 Benchmarking & Competitive Analysis 4.3 Knowledge Management & Best Practice Transfer

6 Chapter 5: Quality Management Systems & Cost of Quality

5.1 Quality Management System (QMS) Frameworks 5.2 Cost of Quality (COQ) Models 5.3 Process Management & Business Process Reengineering 5.4 Management Theories, Styles & Emotional Intelligence 5.5 Communication Skills for Quality Managers 5.6 Project Management for Quality Leaders

7 Chapter 6: Quality Improvement Frameworks & Auditing

6.1 Continuous Improvement Frameworks 6.2 Quality Auditing & Compliance Management 6.3 Measurement Systems & Key Performance Indicators

8 Chapter 7: Seven Basic & Management Quality Tools

7.1 The Seven Basic Quality Tools 7.2 The Seven Management & Planning Tools

9 Chapter 8: Advanced Problem Solving & Risk Analysis

8.1 Failure Mode and Effects Analysis (FMEA) 8.2 Quality Function Deployment (QFD) & House of Quality 8.3 Statistical Process Control & Process Capability 8.4 Lean Tools & Theory of Constraints 8.5 Sampling, MSA & Reliability Terminology

10 Chapter 9: Customer-Focused Organizations & VOC

9.1 Voice of the Customer & Kano Model 9.2 Customer Satisfaction, Loyalty & NPS Metrics 9.3 Customer Relationship Management & Complaint Resolution 9.4 Internal Customers, Segmentation & Multi-Customer Management

11 Chapter 10: Supply Chain Quality Management

10.1 Supplier Qualification, Selection & Evaluation 10.2 Supply Chain Risk Management & Logistics Alignment 10.3 Supplier Audits & Scorecard Systems 10.4 Supplier Risk Management & Communications

12 Chapter 11: Training and Workforce Development

11.1 Training Needs Analysis & Competency Modeling 11.2 Adult Learning Principles & Instructional Design 11.3 Training Program Evaluation & Effectiveness Tracking

8.3 Statistical Process Control & Process Capability

Key Takeaways Explain This Topic AI

SPC was pioneered by Walter A. Shewhart at Bell Labs in 1924 to distinguish between common cause (random) and special cause (assignable) variation.

Variable control charts monitor continuous measurement data (Xbar-R, Xbar-S, I-MR), whereas attribute charts monitor discrete counts (p, np, c, u).

Control limits (UCL/LCL = Mean +/- 3 sigma) reflect the Voice of the Process and are statistically independent of Specification Limits (USL/LSL).

Process capability indices (Cp and Cpk) measure short-term performance under statistical control; Cpk = 1.33 is the standard 4-sigma benchmark.

Process performance indices (Pp and Ppk) evaluate long-term overall process capability including special causes, utilizing overall sample standard deviation (s_total).

Last updated: July 2026

8.3 Statistical Process Control & Process Capability

Exam Tip: ASQ CMQ/OE exam candidates must master the selection rules for control charts (e.g., choosing between c-chart and u-chart based on constant vs. variable sample sizes), calculating Cp and Cpk from given process parameters, and recognizing that specification limits must NEVER be placed on a control chart .

Foundations of Statistical Process Control (SPC)

Statistical Process Control (SPC) is an analytical methodology for monitoring, controlling, and improving a process using statistical techniques. SPC provides real-time feedback to distinguish between expected natural variation and out-of-control signals demanding intervention.

SPC was created in 1924 by Dr. Walter A. Shewhart at Bell Telephone Laboratories, who developed the control chart to eliminate waste in manufacturing telephone equipment. Dr. W. Edwards Deming subsequently popularized Shewhart's concepts in Japan after World War II, establishing statistical control as a cornerstone of modern quality management.

Common Cause vs. Special Cause Variation

A fundamental principle of SPC is categorizing process variation into two distinct sources:

Common Cause Variation (Chance Causes): Natural, inherent noise within a stable system. It is predictable within statistical bounds (Mean +/- 3 Standard Deviations). Individual operators cannot eliminate common cause variation; improving a process affected only by common cause requires management action to redesign the system, upgrade equipment, or change raw materials.

Special Cause Variation (Assignable Causes): Unnatural, erratic signals caused by specific external factors (e.g., broken tool bit, batch of raw material out of spec, operator error, power surge). Special causes make a process statistically unstable. Shop-floor operators must detect, investigate, and correct special causes immediately.

Variable vs. Attribute Control Charts

Selecting the correct control chart depends on the data type (variable continuous vs. attribute discrete) and the subgroup sample size (n) .

  • Data Type Control Chart Monitored Metric Subgroup Sample Size (n) Key Operational Application
  • Variable Xbar - R Chart Subgroup Mean & Range Small (2 <= n <= 8) Precision machining shaft diameters, fills

Variable Xbar - S Chart Subgroup Mean & Std Dev Moderate to Large (n >= 9) Automated high-speed stamping, electronics

Variable I - MR Chart Individual & Moving Range Single Measurement (n = 1) Chemical batch tanks, destructive tensile testing

Attribute p Chart Fraction Nonconforming Variable or Constant (n) Proportion of defective circuit boards per batch

Attribute np Chart Number Nonconforming Constant Subgroup (n) Total count of rejected assemblies per 100 units

Attribute c Chart Count of Nonconformities Constant Area / Opportunity Total scratches on a fixed 1 m^2 glass panel

  • Attribute u Chart Defects Per Unit Variable Area / Opportunity Scratches per m^2 when inspection area varies
  • Selection Rules for Attribute Charts

Use p-chart when evaluating defective items (pass/fail, binary unit classification) when sample size n varies.

Use np-chart for defective items when sample size n is held strictly constant.

Use c-chart when counting defects (nonconformities) on a single unit or fixed inspection area (n = 1 constant unit).

Use u-chart when counting defects (nonconformities) across units where the sample size or inspection area varies.

Control Limits vs. Specification Limits

Confusing control limits with specification limits is a major source of process mismanagement and a common pitfall on ASQ exams. +-------------------------------------------------------------------------+

  • | VOICE OF THE PROCESS (VOP) |
  • | Control Limits (UCL / LCL): Calculated from Process Data (Mean +/- 3SD) |
  • +-------------------------------------------------------------------------+
  • VS
  • +-------------------------------------------------------------------------+
  • | VOICE OF THE CUSTOMER (VOC) |
  • | Specification Limits (USL / LSL): Defined by Engineering & Customers |
  • +-------------------------------------------------------------------------+
  • Upper and Lower Control Limits (UCL / LCL)

Control limits represent the Voice of the Process (VOP) .

Calculated purely from historical process sample data using formulas such as Xbar +/- 3 * Sigma_Xbar or Xbar +/- A2 * Rbar.

Plotted on control charts to evaluate process stability over time.

Upper and Lower Specification Limits (USL / LSL)

Specification limits represent the Voice of the Customer (VOC) .

Defined externally by design engineers, customer drawings, or regulatory requirements.

Control limits and specification limits are statistically independent . A process can be in statistical control (all points within control limits) while simultaneously producing 100% defective parts outside specification limits!

CRITICAL RULE: Specification limits must NEVER be drawn on an item control chart.

Process Capability Indices: Cp and Cpk

Process capability metrics quantify how well a statistically stable process meets customer engineering specifications. Capability indices require the process to be in a state of statistical control and assume a normal distribution of continuous measurement data.

1. Process Capability Ratio (Cp)

Cp measures the potential capability of the process assuming the process mean is perfectly centered between the specification limits. It compares the allowable specification width to the actual natural process spread (6 * sigma_within): C p = USL − LSL 6 σ w i t h i n Cp = \frac{\text{USL} - \text{LSL}}{6\sigma_{within}} Cp = 6 σ w i t hin ​ USL − LSL ​

Where short-term standard deviation sigma_within is estimated from subgroup ranges or standard deviations: σ w i t h i n = R ˉ d 2 or σ w i t h i n = S ˉ c 4 \sigma_{within} = \frac{\bar{R}}{d_2} \quad \text{or} \quad \sigma_{within} = \frac{\bar{S}}{c_4} σ w i t hin ​ = d 2 ​ R ˉ ​ or σ w i t hin ​ = c 4 ​ S ˉ ​

Note: Cp measures process spread relative to specification tolerance, but it completely ignores where the process mean (mu) is located.

2. Process Capability Index (Cpk)

Cpk measures the actual capability of the process by accounting for process off-centering relative to specification boundaries: C p k = min ⁡ ( USL − μ 3 σ w i t h i n , μ − LSL 3 σ w i t h i n ) Cpk = \min \left( \frac{\text{USL} - \mu}{3\sigma_{within}}, \, \frac{\mu - \text{LSL}}{3\sigma_{within}} \right) Cp k = min ( 3 σ w i t hin ​ USL − μ ​ , 3 σ w i t hin ​ μ − LSL ​ )

Interpreting Capability Benchmark Values

Cpk < 1.00: Process is incapable . The natural process spread exceeds tolerance limits, producing non-conforming product (> 2,700 DPMO).

Cpk = 1.00: Process is barely capable (3-sigma capability). 0.27% defects (2,700 DPMO) for two-sided centered specifications.

Cpk = 1.33: Minimum standard industry benchmark for mature processes (4-sigma capability, approx. 63 DPMO).

Cpk = 1.67: Automotive and aerospace standard for new critical processes (5-sigma capability, approx. 0.57 DPMO).

Cpk = 2.00: Six Sigma capability (3.4 DPMO considering a 1.5-sigma long-term mean shift).

When a process is perfectly centered (mu = (USL + LSL) / 2), Cp = Cpk. As the process mean drifts off-center, Cpk decreases while Cp remains constant. Therefore, Cpk is always less than or equal to Cp (Cpk <= Cp).

Process Performance Indices: Pp and Ppk

While Cp and Cpk evaluate short-term potential using within-subgroup variation (sigma_within), Process Performance Indices (Pp and Ppk) evaluate overall long-term performance across extended operating periods, incorporating both common causes and potential special cause shifts.

Performance formulas substitute overall sample standard deviation (s_total) calculated across all individual observations N: s t o t a l = ∑ i = 1 N ( X i − X ˉ ˉ ) 2 N − 1 s_{total} = \sqrt{\frac{\sum_{i=1}^{N} (X_i - \bar{\bar{X}})^2}{N - 1}} s t o t a l ​ = N − 1 ∑ i = 1 N ​ ( X i ​ − X ˉ ˉ ) 2 ​ ​ P p = USL − LSL 6 s t o t a l Pp = \frac{\text{USL} - \text{LSL}}{6 s_{total}} Pp = 6 s t o t a l ​ USL − LSL ​ P p k = min ⁡ ( USL − X ˉ ˉ 3 s t o t a l , X ˉ ˉ − LSL 3 s t o t a l ) Ppk = \min \left( \frac{\text{USL} - \bar{\bar{X}}}{3 s_{total}}, \, \frac{\bar{\bar{X}} - \text{LSL}}{3 s_{total}} \right) Pp k = min ( 3 s t o t a l ​ USL − X ˉ ˉ ​ , 3 s t o t a l ​ X ˉ ˉ − LSL ​ )

Comparison: Capability vs. Performance

Cpk approx. Ppk: Indicates a highly stable process under statistical control with no significant shifts between subgroups over time.

Cpk >> Ppk: Indicates significant process drift, batch-to-batch variation, or uncorrected special cause events across subgroups.

Process Capability Index Benchmarks vs DPMO (Defects Per Million Opportunities)