In the world of process improvement and quality management, understanding variation is fundamental to making informed decisions. Control charts serve as one of the most powerful statistical tools available to organizations implementing Lean Six Sigma methodologies. These visual instruments help teams distinguish between normal process fluctuations and genuine problems that require intervention. During the Measure phase of a Lean Six Sigma project, control charts become indispensable for establishing baselines and identifying opportunities for improvement.
What Are Control Charts?
Control charts, also known as Shewhart charts or process behavior charts, are graphical representations of process data over time. Developed by Walter Shewhart in the 1920s, these charts plot data points in chronological order with statistically determined upper and lower control limits. The centerline typically represents the process mean or median, while the control limits define the boundaries of expected process variation. You might also enjoy reading about Sampling Methods in Six Sigma: Understanding Random, Stratified, and Systematic Sampling Techniques.
The primary purpose of a control chart is to monitor process stability and distinguish between two types of variation: common cause variation and special cause variation. This distinction forms the cornerstone of effective process management and improvement initiatives within the lean six sigma framework. You might also enjoy reading about Attribute Agreement Analysis: A Complete Guide to Measuring Consistency in Go/No-Go Decisions.
Understanding Process Variation
Before diving deeper into control charts, it is essential to understand the concept of variation itself. No process operates with absolute consistency. Every manufacturing line, service delivery system, or business process experiences some degree of fluctuation in its outputs. Recognizing and properly categorizing this variation determines whether corrective action is necessary or counterproductive. You might also enjoy reading about Measure Phase Timeline: How Long Should Data Collection Really Take in Lean Six Sigma Projects.
Common Cause Variation
Common cause variation, also called random variation or noise, represents the natural fluctuation inherent in any process. These variations result from numerous small factors that are always present and create a predictable pattern of dispersion. When only common cause variation exists, a process is considered stable or “in control.” Examples include slight differences in raw material properties, minor environmental temperature changes, or normal human performance variability.
Attempting to adjust a process experiencing only common cause variation often creates more problems than it solves, a phenomenon known as tampering. This counterproductive intervention increases overall variation and destabilizes the process.
Special Cause Variation
Special cause variation, also known as assignable cause variation, arises from specific, identifiable factors that are not part of the normal process operation. These causes produce unusual patterns or points that fall outside expected boundaries. Examples include equipment malfunction, operator error, material defects, or sudden environmental changes.
When special cause variation appears, immediate investigation and corrective action become necessary. Identifying and eliminating these exceptional sources of variation is a primary objective during the recognize phase of problem identification within quality improvement projects.
The Role of Control Charts in the Measure Phase
Within the DMAIC (Define, Measure, Analyze, Improve, Control) framework of lean six sigma, the Measure phase focuses on quantifying current process performance and establishing a baseline for comparison. Control charts play several critical roles during this phase.
Establishing Process Stability
Before analyzing process capability or making improvements, teams must first determine whether the process is stable. A stable process exhibits only common cause variation, making its behavior predictable. Control charts provide the visual and statistical evidence needed to assess stability. Without this stability assessment, any capability analysis or improvement effort may be built on a foundation of unreliable data.
Creating a Baseline for Improvement
Control charts document the current state of a process, creating a quantitative baseline against which future improvements can be measured. This baseline becomes the reference point for demonstrating project success and calculating return on investment. During the recognize phase of problem identification, this baseline helps teams understand the magnitude of issues and prioritize improvement opportunities.
Identifying Improvement Opportunities
By revealing special cause variation, control charts highlight specific instances or periods when processes deviate from normal operation. These deviations represent opportunities for investigation and improvement. Teams can examine what was different during these periods and either eliminate negative special causes or incorporate positive ones into standard operations.
Types of Control Charts
Different types of control charts suit different types of data and process characteristics. Selecting the appropriate chart type is crucial for accurate analysis.
Variable Data Charts
Variable data consists of continuous measurements such as time, temperature, length, or weight. The most common variable data control charts include:
- X-bar and R Charts: These paired charts monitor both the process mean (X-bar) and range (R) for subgroups of data, making them suitable for tracking central tendency and dispersion simultaneously.
- X-bar and S Charts: Similar to X-bar and R charts but using standard deviation (S) instead of range, these charts are preferred for larger subgroup sizes.
- Individual and Moving Range (I-MR) Charts: These charts work with individual measurements rather than subgroups, ideal when sampling is expensive or destructive, or when measurements occur infrequently.
Attribute Data Charts
Attribute data represents counts or classifications such as defective versus non-defective, or the number of errors. Common attribute control charts include:
- P Charts: Monitor the proportion of defective units in samples of varying sizes.
- NP Charts: Track the number of defective units when sample size remains constant.
- C Charts: Count the number of defects per unit when the sample size or area of opportunity stays constant.
- U Charts: Monitor the rate of defects per unit when sample size or area varies.
Interpreting Control Charts
Reading control charts requires understanding specific patterns and rules that indicate special cause variation. While a point outside control limits clearly signals special cause variation, other patterns also warrant investigation.
Western Electric Rules
The Western Electric rules, also known as the Nelson rules, provide systematic criteria for detecting special cause variation. These include:
- One point beyond the control limits
- Two out of three consecutive points in Zone A (outer third between centerline and control limit) or beyond
- Four out of five consecutive points in Zone B (middle third) or beyond
- Eight consecutive points on one side of the centerline
- Six consecutive points steadily increasing or decreasing (a trend)
- Fifteen consecutive points within Zone C (closest third to centerline)
- Fourteen consecutive points alternating up and down
- Eight consecutive points avoiding Zone C
Applying these rules helps teams recognize phase patterns that may not be immediately obvious but still indicate process instability requiring investigation.
Implementing Control Charts Effectively
Successfully implementing control charts within lean six sigma projects requires following several best practices.
Collect Sufficient Data
Gather enough data points to calculate meaningful control limits, typically a minimum of 20 to 25 subgroups. Insufficient data produces unreliable limits that may lead to incorrect conclusions about process stability.
Ensure Measurement System Accuracy
Before charting data, conduct measurement system analysis to verify that measurement instruments and procedures provide accurate, precise, and repeatable results. Measurement error can mask or create false signals of special cause variation.
Update Limits Appropriately
After removing special causes and stabilizing a process, recalculate control limits to reflect the improved process capability. Similarly, update limits when fundamental process changes occur intentionally through improvement initiatives.
Respond to Signals Promptly
Establish clear protocols for investigating and responding to out-of-control signals. Delayed responses allow problems to persist and potentially worsen, undermining the value of control chart monitoring.
Conclusion
Control charts represent an essential tool for understanding and managing process variation during the Measure phase of lean six sigma projects. By distinguishing between common cause and special cause variation, these charts enable teams to make data-driven decisions about when to intervene and when to leave processes alone. Mastering control chart basics empowers organizations to establish accurate baselines, recognize improvement opportunities, and build a foundation for sustainable process enhancement. As organizations continue their quality improvement journeys, control charts remain indispensable for maintaining process control and driving operational excellence.
