CSSGB Control Phase: SPC Charts, Control Plans, and Sustaining Improvements

Control Phase Overview: The Final 15% of Your CSSGB Exam

The Control Phase is where Six Sigma projects either become permanent fixtures of organizational excellence or slowly revert to their pre-improvement state. As the final phase of the DMAIC cycle, Control accounts for 15% of the ASQ Certified Six Sigma Green Belt exam, making it the lightest-weighted domain yet one of the most conceptually distinct. While the earlier phases focus on measuring problems and generating solutions, the Control Phase is entirely about locking in gains and ensuring your process improvements survive long after the project team disbands.

If you have been studying the DMAIC framework from the beginning, you already understand how the Define Phase establishes project scope and charter requirements and how the Improve Phase uses DOE, lean tools, and kaizen concepts to implement solutions. The Control Phase builds on all of that work by answering one critical question: how do we make sure these improvements stick?

The ASQ Body of Knowledge for the Control Phase organizes content into three primary areas: Statistical Process Control (SPC), Control Plans, and Sustaining Improvements. Each area demands both conceptual understanding and the ability to apply tools in real-world scenarios. For a complete breakdown of all six DMAIC domains, see our guide on the ASQ Six Sigma Green Belt Body of Knowledge.

Statistical Process Control (SPC) Fundamentals

Statistical Process Control is the backbone of the Control Phase. SPC uses statistical methods to monitor and control a process, ensuring it operates at its full potential while producing conforming product. At its core, SPC distinguishes between two types of variation that every Green Belt must understand cold for the exam.

Common Cause vs. Special Cause Variation

Common cause variation (also called chance cause or random variation) is the natural, inherent variability built into every process. It is predictable, stable, and consistent over time. Think of it as the background noise of a process. A process exhibiting only common cause variation is said to be in statistical control or stable.

Special cause variation (also called assignable cause variation) is variation that arises from specific, identifiable sources outside the normal process behavior. A machine malfunction, a new operator, a batch of defective raw material — these are all special causes. When special cause variation is present, the process is out of statistical control and unpredictable.

The Purpose of Control Charts

Control charts are the primary SPC tool. Developed by Walter Shewhart in the 1920s, they plot process data over time against statistically derived upper and lower control limits (UCL and LCL). The center line represents the process average. Control limits are typically set at ±3 standard deviations from the center line, capturing approximately 99.73% of natural process variation.

Control charts serve three functions that appear frequently on the exam:

• Monitoring — Detecting shifts, trends, and out-of-control conditions in real time
• Diagnosis — Distinguishing between common and special cause variation
• Decision-making — Determining when to investigate a process and when to leave it alone

Control Chart Types You Must Know for the Exam

The CSSGB exam requires you to know when to use each type of control chart. The two major categories are variables charts (for continuous, measurable data) and attributes charts (for count or classification data). Each category contains specific chart types suited to different data situations.

Chart Type	Data Type	What It Monitors	When to Use
X̄-R (X-bar and Range)	Variables (continuous)	Process mean and variability	Subgroup size 2–10; most common variables chart
X̄-S (X-bar and Std Dev)	Variables (continuous)	Process mean and variability	Subgroup size >10; more precise than R chart
I-MR (Individuals and Moving Range)	Variables (continuous)	Individual measurements and variability	Subgroup size = 1; slow production or destructive testing
p chart	Attributes (defective)	Proportion defective	Variable or constant sample size; binary classification
np chart	Attributes (defective)	Number defective	Constant sample size only; binary classification
c chart	Attributes (defects)	Count of defects per unit	Constant sample size; multiple defects possible per unit
u chart	Attributes (defects)	Defects per unit	Variable sample size; multiple defects possible per unit

Variables Charts in Detail

The X̄-R chart is the workhorse of SPC. It pairs two charts: the X̄ (X-bar) chart monitors the process mean across subgroups, while the R (Range) chart monitors within-subgroup variability. When you collect rational subgroups of 2–10 measurements, this is your default choice. For subgroups larger than 10, the X̄-S chart replaces the range chart with a standard deviation chart for improved accuracy.

The I-MR chart (Individuals and Moving Range) handles situations where you can only collect one measurement at a time. This is common in chemical batch processes, expensive testing, or slow production environments. The moving range is calculated as the absolute difference between consecutive individual measurements.

Attributes Charts in Detail

Attributes charts deal with data that can be counted or classified. The critical distinction for the exam is between defective items (a binary pass/fail classification for the entire unit) and defects (a count of individual nonconformities, where a single unit can have multiple defects).

The p chart and np chart both track defectives. Use the p chart when sample sizes vary between subgroups because it normalizes to a proportion. Use the np chart when sample sizes are constant because working with raw counts is simpler. The c chart and u chart both track defects. The c chart requires constant sample sizes, while the u chart handles variable sample sizes by expressing defects per unit.

Interpreting Control Charts: Rules and Out-of-Control Signals

Knowing how to build a control chart is only half the battle. The CSSGB exam places heavy emphasis on your ability to interpret control chart patterns and identify out-of-control conditions. The Western Electric rules (also called the Nelson rules or Shewhart rules, depending on your reference material) define specific patterns that signal non-random behavior.

The Eight Nelson Rules

The remaining rules involve cyclical patterns, alternating points, and other non-random configurations. For the CSSGB exam, rules 1 through 4 are the most frequently tested. Make sure you can recognize each pattern visually and explain the likely cause.

Remember that since the CSSGB is an open-book exam, you can bring reference material containing these rules and chart constants. For strategies on organizing your reference materials, see our ASQ CSSGB Exam Day Tips: Open-Book Strategies and Reference Material Guide.

Control Plans: Documenting and Standardizing Improvements

A control plan is a living document that describes the system for controlling a process after improvements have been implemented. It is the bridge between the project team's work and the process operators who will run the improved process day after day. The exam tests your understanding of what a control plan contains and how it fits into the overall DMAIC closure process.

Elements of an Effective Control Plan

A comprehensive control plan typically includes the following elements:

• Process steps — Each critical step in the process that requires monitoring
• Key input variables (X's) — The process inputs identified during the Analyze and Improve phases
• Key output variables (Y's) — The process outputs and CTQ characteristics
• Specifications and tolerances — Target values and acceptable ranges for each variable
• Measurement method — How each variable will be measured (instrument, gage, test method)
• Sample size and frequency — How many units to measure and how often
• Control method — The specific SPC chart, checklist, or monitoring technique in use
• Reaction plan — Exact steps to follow when an out-of-control condition is detected
• Responsible person — Who owns each monitoring activity and each response action

Standard Operating Procedures (SOPs)

Control plans are supported by Standard Operating Procedures, detailed step-by-step instructions for performing process tasks in a consistent manner. SOPs ensure that the improved process methods become the new standard rather than a suggestion. Effective SOPs should be written clearly enough that a trained operator can follow them without interpretation. They should specify exact settings, sequences, and criteria rather than vague directives.

Response Plans and Escalation

When a control chart signals an out-of-control condition, the response plan dictates the actions to take. A well-designed response plan includes immediate containment actions, root cause investigation steps, corrective action requirements, and escalation paths if the issue cannot be resolved at the operator level. The response plan essentially automates the decision-making process so that gains are protected even when the original project team is not available.

Sustaining Improvements: Lessons Learned and Knowledge Transfer

One of the most underappreciated aspects of the Control Phase is the human side of sustaining improvements. Statistical tools and documentation only work if people follow them, and people only follow them if they understand the reasons behind the changes and feel ownership of the new process.

Training and Knowledge Transfer

Effective training ensures that all stakeholders — operators, supervisors, quality engineers, and management — understand the improved process and their role in maintaining it. Training should cover not only the how (new procedures, SPC charts, response plans) but also the why (what problems existed before, what improvements were achieved, what could go wrong if controls lapse).

The exam may ask about methods for ensuring knowledge transfer, including documentation handoff, train-the-trainer programs, visual workplace displays, and periodic audits. The key concept is that knowledge must be embedded in the system (via SOPs, control plans, and visual controls) rather than residing solely in the project team's heads.

Lessons Learned

Lessons learned documentation captures what worked well, what could be improved, and what was discovered during the project. This knowledge benefits future projects and prevents organizations from repeating mistakes. A formal lessons learned session typically occurs during project closure and feeds into the organization's knowledge management system.

Project Closure and Handoff

The Control Phase concludes with formal project closure, which includes validating that improvements are sustained over a predetermined monitoring period, transferring ownership of the control plan to the process owner, updating process documentation and FMEA, presenting final results to leadership, and archiving project documentation for future reference.

Lean Controls: 5S, Visual Management, and Mistake-Proofing

While SPC provides the statistical framework for process control, lean tools provide practical, shop-floor-level controls that complement charting and documentation. The CSSGB exam includes several lean control mechanisms within the Control Phase domain.

5S as a Control Mechanism

The 5S methodology (Sort, Set in Order, Shine, Standardize, Sustain) is a workplace organization system that serves as a visual control mechanism. From a Control Phase perspective, Standardize and Sustain are the critical S's. Standardize establishes the procedures and visual cues that define the "correct" state of the workspace. Sustain creates the audit and accountability structure that keeps standards from degrading over time.

Poka-Yoke (Mistake-Proofing)

Poka-yoke devices are mistake-proofing mechanisms that either prevent errors from occurring or immediately detect them when they do occur. Examples include asymmetric connectors that can only be inserted one way, sensors that detect missing components before assembly proceeds, and software validations that prevent invalid data entry. Poka-yoke is one of the most powerful control mechanisms because it removes the reliance on human vigilance.

Control Mechanism	Effectiveness Level	Human Dependence	Example
Poka-Yoke (Prevention)	Highest	None — error is impossible	Asymmetric connector that only fits one way
Poka-Yoke (Detection)	High	Minimal — immediate alert	Weight sensor that stops the line if part is missing
SPC with Reaction Plan	Moderate-High	Requires trained operator response	Control chart with documented escalation steps
Visual Controls	Moderate	Requires operator attention	Color-coded bins, floor markings, shadow boards
SOP / Work Instructions	Moderate-Low	Relies on operator compliance	Written procedure posted at workstation
Training Only	Lowest	Entirely dependent on memory	Classroom training without supporting documentation

Visual Management

Visual management systems make process status, performance, and standards visible at a glance. Examples include performance dashboards displayed on screens or whiteboards, color-coded signals for process status (green/yellow/red), shadow boards for tools so missing items are immediately obvious, and floor markings that define work zones, walkways, and material locations. Visual controls reduce reliance on memory and make deviations from standard immediately apparent.

Exam Strategy for Control Phase Questions

With approximately 16 scored questions dedicated to the Control Phase, strategic preparation is essential. Here is how to maximize your score on this domain.

For a comprehensive approach to tackling all six domains, our complete study guide for the ASQ Six Sigma Green Belt exam provides a structured roadmap that integrates Control Phase preparation with the other DMAIC phases.

Sample Control Phase Questions

Testing yourself with practice questions is one of the most effective ways to prepare for the CSSGB exam. Below are examples of the types of questions you may encounter on Control Phase topics.

Question 1: A quality engineer is monitoring the number of paint defects on automobile body panels. Each panel can have multiple defects, and the number of panels inspected varies each shift. Which control chart is most appropriate?

Answer: The u chart. The data involves defects (not defectives), multiple defects per unit are possible, and the sample size varies — all of which point to the u chart.

Question 2: On an X̄-R control chart, eight consecutive points fall above the center line on the X̄ chart. What does this indicate?

Answer: The process is out of statistical control. A run of eight consecutive points on one side of the center line violates the run rule and signals a likely shift in the process mean due to a special cause.

Question 3: What is the primary purpose of a reaction plan within a control plan?

Answer: To define the specific actions to be taken when an out-of-control condition or nonconformance is detected, including containment, investigation, and corrective action steps.

For more practice with exam-style questions across all six domains, visit our CSSGB practice test site and try our free CSSGB sample problems.

Frequently Asked Questions