Study guide
9.2 Common Cause vs Special Cause Variation
ASQ Certified Six Sigma Yellow Belt CSSYB
Track your progress as you read
Finished sections get a check mark.
1 Introduction & Exam Overview
About the ASQ CSSYB Exam How to Use This Study Guide
2 Chapter 1: Six Sigma & Lean Fundamentals
1.1 Six Sigma Foundations, Principles & History 1.2 Lean Foundations, Principles & The 8 Wastes 1.3 Six Sigma Roles & Responsibilities
3 Chapter 2: Team Basics & Dynamics
2.1 Team Types & Stages of Team Evolution 2.2 Team Decision-Making & Idea Generation Tools 2.3 Communication Methods & Meeting Management
4 Chapter 3: Quality Tools & Six Sigma Metrics
3.1 The 7 Basic Quality Tools 3.2 Six Sigma Process Performance Metrics & COPQ
5 Chapter 4: Define Phase — Project Identification & Scope
4.1 Voice of the Customer (VOC) & Critical-to-Quality (CTQ) 4.2 Project Selection & Stakeholder Analysis 4.3 Process Definition with SIPOC & Supply Chain Basics
6 Chapter 5: Define Phase — Project Management Basics
5.1 The Project Charter & Problem Statements 5.2 Project Planning: WBS, Gantt Charts & Communication Plans 5.3 Project Management Tools & Tollgate Reviews
7 Chapter 6: Measure Phase — Data Collection & Statistics
6.1 Basic Statistics: Central Tendency & Dispersion 6.2 Data Types & Classification 6.3 Data Collection Plans, Operational Definitions & Sampling 6.4 Data Collection Techniques: Check Sheets, Checklists, Surveys & Interviews
8 Chapter 7: Measure Phase — Measurement System Analysis
7.1 Measurement System Fundamentals: Precision, Accuracy & Bias 7.2 Gauge Repeatability & Reproducibility (GR&R)
9 Chapter 8: Analyze Phase — Process & Root Cause Analysis
8.1 Process Analysis: 5S & Value Analysis 8.2 Failure Mode and Effects Analysis (FMEA) 8.3 Root Cause Analysis: 5 Whys, Process Mapping, 8D, Force-Field & Matrix Charts
10 Chapter 9: Analyze Phase — Corrective Action & Variation
9.1 Corrective & Preventive Action Processes 9.2 Common Cause vs Special Cause Variation 9.3 Basic Distribution Types: Normal, Binomial & Poisson
11 Chapter 10: Analyze Phase — Relationships & Hypothesis Testing
10.1 Correlation & Scatter Plots 10.2 Simple Linear Regression & Prediction 10.3 Hypothesis Testing Concepts: Hypotheses, Errors & P-Values
12 Chapter 11: Improve Phase — Implementation Techniques
11.1 Kaizen & Kaizen Blitz Events 11.2 Plan-Do-Check-Act (PDCA) & Solution Selection 11.3 Cost-Benefit Analysis & Financial Validation
13 Chapter 12: Control Phase — Sustaining Improvements
12.1 Control Plans & Process Monitoring 12.2 Statistical Process Control: X-bar & R Charts 12.3 Document Control, Standard Operating Procedures & Project Hand-off
9.2 Common Cause vs Special Cause Variation
Key Takeaways Explain This Topic AI
Common cause variation is the natural, inherent background noise of a stable system governed by random chance.
Special cause variation represents assignable, intermittent disturbances arising from specific external disruptions outside baseline design.
Process stability must precede process capability assessment; calculating Cp or Cpk on an unstable process produces statistically invalid results.
Tampering—making ad-hoc adjustments to a stable process displaying only common cause variation—mathematically inflates process variance.
Walter Shewhart and W. Edwards Deming demonstrated that common cause variation requires system redesign by management, whereas frontline teams resolve special causes.
Last updated: September 2026
Common Cause vs Special Cause Variation
Quick Answer: In statistical quality control, process variation divides into two fundamental categories: common cause (inherent, predictable background noise) and special cause (assignable, intermittent disturbances). Walter Shewhart and W. Edwards Deming demonstrated that confusing these two variations leads to catastrophic management errors. A process must attain statistical stability (exhibiting only common causes) before process capability ( C p , C p k C_p, C_{pk} C p , C p k ) can be calculated. Making ad-hoc adjustments to a stable process—known as tampering—mathematically increases overall variance. Independent CSSYB study guide by ExamFinch.
The Foundations of Variation Theory: Shewhart and Deming
Every operational process—whether manufacturing automotive camshafts, processing commercial loan applications, or dispensing pharmaceuticals—exhibits variation. No two products, transactions, or service interactions are perfectly identical. In the 1920s at Bell Telephone Laboratories, physicist Walter A. Shewhart recognized that failing to understand the nature of variation creates massive operational waste.
Shewhart, whose work was later expanded globally by W. Edwards Deming, proved that variation arises from two completely distinct sources: common causes and special causes . Deming famously asserted that misinterpreting variation—attributing a common cause to a special cause or vice versa—represents one of the most costly mistakes in management. For Yellow Belts participating in DMAIC projects, correctly diagnosing variation is the cornerstone of the Analyze phase.
Common Cause Variation (Chance / Inherent Variation)
Common cause variation refers to the natural, random background noise inherent to the established design of a system. It represents the cumulative effect of countless minor, unidentifiable, and unavoidable variables acting simultaneously.
Key Characteristics
Inherent to the System: Common cause variation is built directly into current machinery tolerances, standard material chemistry, ambient workplace humidity, and standard operating procedures.
Statistically Stable and Predictable: A process influenced solely by common cause variation is in statistical process control . While individual data points cannot be predicted in advance, the distribution of outputs over time is predictable within established statistical bounds (the Upper and Lower Control Limits, ± 3 σ \pm 3\sigma ± 3 σ ).
Examples: Minor electrical line voltage fluctuations within utility tolerances; micro-variations in vendor resin pellet density; normal slight shifts in operator hand position during assembly; seasonal ambient temperature swings within an HVAC-controlled facility.
Action Required for Common Causes
Common cause variation cannot be resolved by counseling frontline operators or making day-to-day machine tweaks. Because common cause variation stems from system design, reducing it requires fundamental process redesign led by management. Deming estimated that 85% to 94% of operational problems stem from common causes belonging to system design, which only management has the authority to alter (such as purchasing higher-precision machinery, switching to premium-grade raw materials, or overhauling workflow layout).
Special Cause Variation (Assignable Variation)
Special cause variation refers to unpredictable, intermittent, and external disruptions that are not part of the standard process design. These disturbances arise from specific, identifiable events.
Key Characteristics
Assignable and External: Special causes represent localized events outside normal operation—such as a broken cutting tool, an uncalibrated measurement instrument, a new operator lacking standard training, a batch of contaminated solvent, or a power outage.
Statistically Unpredictable: A process subject to special causes is unstable or out of control . Its future performance cannot be statistically predicted, and output parameters drift, shift, or spike.
Control Chart Manifestations: Special causes are flagged on Shewhart control charts by points falling outside the three-sigma control limits, or by non-random patterns such as a run of consecutive points on one side of the center line (seven points in the AIAG rule set, eight in the Western Electric rules, nine in the Nelson rules), six consecutive points steadily increasing or decreasing (a trend), or fourteen points alternating up and down.
Action Required for Special Causes
Special cause variation demands immediate frontline intervention . Frontline operators and Yellow Belts must identify the specific assignable cause, contain suspect output, isolate the root factor at the gemba, and remove the disturbance to restore the process to statistical stability.
Direct Comparison: Common Cause vs. Special Cause Variation
Attribute Common Cause Variation Special Cause Variation
Synonyms Chance cause, inherent variation, background noise, random variation Assignable cause, non-random variation, external disturbance
Source Entire system design, environment, and standard operating baseline Specific, localized events, disruptions, or parameter changes
Predictability Predictable within fixed statistical control limits ( ± 3 σ \pm 3\sigma ± 3 σ ) Unpredictable in timing, magnitude, and direction
Process State In Statistical Control (Stable) Out of Control (Unstable)
Control Chart Signature Points vary randomly within control limits without non-random patterns Points outside control limits, runs, trends, or non-random clustering
Responsibility Management (requires fundamental system/process redesign) Frontline personnel & Yellow Belts (local root cause isolation & removal)
- Management Error Treating as special cause leads to Tampering Treating as common cause leads to Ignored Crises
- Process Stability vs. Process Capability: The Precedence Rule
One of the most vital principles tested in Six Sigma is the strict sequential relationship between Process Stability and Process Capability .
1. Process Stability (Voice of the Process)
Process stability measures whether the process is predictable over time. It is determined exclusively by comparing process data against statistical control limits calculated from internal process variance ( X ˉ ± 3 σ \bar{X} \pm 3\sigma X ˉ ± 3 σ ). A stable process exhibits only common cause variation.
2. Process Capability (Voice of the Customer)
Process capability ( C p , C p k C_p, C_{pk} C p , C p k ) evaluates whether the process output meets external customer tolerance specifications (Upper and Lower Specification Limits, USL and LSL).
The Golden Rule of Quality Engineering
A process MUST achieve statistical stability before process capability can be meaningfully evaluated. Statistical Stability (Control) ⟶ Process Capability Assessment ( C p , C p k ) \text{Statistical Stability (Control)} \longrightarrow \text{Process Capability Assessment } (C_p, C_{pk}) Statistical Stability (Control) ⟶ Process Capability Assessment ( C p , C p k )
If a process is unstable (subject to special cause variation), its mean and standard deviation constantly fluctuate. Calculating C p C_p C p or C p k C_{pk} C p k on an unstable process yields ephemeral, mathematically meaningless numbers. A Yellow Belt must first identify and eliminate all assignable causes to establish a stable baseline before assessing capability.
Process Tampering and Deming's Funnel Experiment
Tampering occurs when an operator or manager treats common cause variation as if it were a special cause by making frequent, ad-hoc adjustments to a stable process in response to individual output measurements.
Deming's Funnel Experiment
To demonstrate the mathematical destruction caused by tampering, Deming used a funnel suspended above a stationary target on a table, dropping marbles through the funnel to observe where they landed:
Rule 1 (No Adjustment): The funnel is left fixed in place over the target. The marbles scatter naturally around the target, forming a stable normal distribution with baseline variance σ 2 \sigma^2 σ 2 . This represents appropriate management of common cause variation.
Rule 2 (Adjust from Last Error): After each drop, the funnel is moved from its current position by an amount equal and opposite to the error of the last marble. Result: The variance doubles ( 2 σ 2 2\sigma^2 2 σ 2 ), and the scatter circle expands significantly.
Rule 3 (Adjust from Target): After each drop, the funnel is reset relative to the original target by an amount equal and opposite to the error of the last marble. Result: The funnel oscillates back and forth with wildly explosive variance, sending marbles flying across the room.
Rule 4 (Set Over Last Drop): The funnel is moved to sit directly over the exact spot where the last marble landed. Result: The process exhibits steady drift, wandering steadily away from the target without bound.
Key Lesson for Yellow Belts
Adjusting a stable machine or workflow after observing a single piece that was slightly high or low invariably increases total process variation. Yellow Belts must leave stable processes alone unless statistical control charts indicate an assignable special cause.