Control Charts Explained: A Complete Guide to SPC for Six Sigma Professionals

Every process varies. Two parts made on the same machine are never exactly identical, and a service delivered by the same team takes slightly different amounts of time on different days. The real question is whether that variation is normal or a sign that something has gone wrong.

That is the question control charts in Six Sigma answer. They show you when to leave a process alone and when to step in.

In this guide, we’ll cover what control charts are, how they fit into statistical process control (SPC), the main types of control charts, and how to read them correctly. Whether you’re preparing for a Six Sigma Green Belt or Black Belt or just want better process data, this guide is for you.

What Are Control Charts in Six Sigma?

A control chart is a time-based graph that tracks how a process behaves. It plots your data points in order and adds three reference lines:

• Center line (CL): the average of the process
• Upper control limit (UCL): the highest value you’d expect from normal variation
• Lower control limit (LCL): the lowest value you’d expect from normal variation

Walter Shewhart developed the idea at Bell Labs in the 1920s. Today it is one of the most important tools in the Control phase of DMAIC, where the goal is to hold the gains an improvement project has made.

A control chart is not a pass/fail test against customer specifications. The control limits come from the process data itself, so the chart shows what the process is doing, not what you want it to do. This distinction confuses many beginners, and we’ll come back to it below.

Understanding Statistical Process Control (SPC)

Statistical process control (SPC) is the practice of using statistical methods to monitor and control a process. Instead of inspecting finished products and sorting good from bad, SPC helps you catch problems while the process is running.

The control chart is the heart of SPC. Other SPC tools include histograms, Pareto charts, and process capability analysis. If you want a refresher on the basics, our guide to the 7 QC tools covers the foundations that sit alongside statistical process control charts, which are the ones that give you real-time signals.

The benefits of using SPC are practical:

• You catch problems early, before defects pile up
• You avoid over-reacting to random noise
• You reduce scrap, rework, and customer complaints
• You make decisions from data rather than gut feeling

Professionals trained through programs like those at Gravitex Genesys learn to apply SPC across manufacturing, healthcare, IT, banking, and logistics, because process variation shows up everywhere.

Process Variation in Six Sigma: Common Cause vs. Special Cause

You can’t use a control chart well without understanding process variation in Six Sigma. There are two kinds.

Common Cause Variation

This is the natural, built-in variation of a stable process. It comes from many small factors: slight material differences, normal temperature shifts, small differences in how people work. A process with only common cause variation is called in control or stable. Its output is predictable, even if the output isn’t perfect.

Special Cause Variation

This is variation from a specific, identifiable source: a worn tool, a bad batch of raw material, a new untrained operator, a power surge. It’s not part of the normal process. A process affected by special causes is out of control and unpredictable.

Why it matters: The two types need different responses. If you chase common cause variation with quick fixes, you often make things worse, a mistake Deming called “tampering.” If you ignore special cause variation, problems keep growing. Control charts tell you which one you’re looking at.

The Anatomy of a Control Chart

Before looking at the types, here is how a typical chart is built:

1. Collect data in time order, using rational subgroups where appropriate
2. Calculate the center line, which is the process mean
3. Calculate the control limits, typically set at ±3 standard deviations (3 sigma) from the mean
4. Plot the points and connect them
5. Interpret the patterns using detection rules

Using three-sigma limits means that if a process is stable, about 99.73% of points should fall inside the limits. A point outside them is unlikely enough by chance that it deserves investigation.

Types of Control Charts

Choosing the right chart is the first real decision in SPC, and it depends on the kind of data you have. Broadly, there are two families of control charts.

Control Charts for Continuous (Variable) Data

Use these when you measure something on a scale: length, weight, temperature, time, or cost.

X-bar and R Chart
Tracks the subgroup average (X-bar) and the range within each subgroup (R). It works well for subgroup sizes of about 2 to 10. It’s the classic chart for manufacturing measurements.

X-bar and S Chart
Similar to X-bar and R, but it uses the standard deviation (S) instead of the range. It’s better for larger subgroups, usually more than 10.

Individuals and Moving Range (I-MR) Chart
Use this when you only have one measurement at a time, such as daily revenue, batch results, or monthly readings. It’s one of the most widely used charts in service industries and transactional processes.

Control Charts for Discrete (Attribute) Data

Use these when you count things: defects, errors, or defective units.

P Chart
Tracks the proportion of defective items in a sample. Sample sizes can vary. Example: the percentage of invoices with errors each week.

NP Chart
Tracks the number of defective items when the sample size is constant.

C Chart
Tracks the count of defects in a constant-size unit. Example: the number of scratches on each panel.

U Chart
Tracks defects per unit when the sample size or area changes. Example: the number of bugs found per 1,000 lines of code.

A Quick Way to Choose

If you’re unsure, I-MR is a safe starting point for continuous data, and the P chart for defective rates.

How to Do Control Chart Analysis: Reading the Signals

Plotting the chart is the easy part. Control chart analysis is about reading what the pattern says. A process is considered to show special cause variation when you see any of the following signals (based on the commonly used Western Electric and Nelson rules):

1. One point beyond the control limits. The clearest signal of all.
2. Eight or more consecutive points on one side of the center line. This suggests a shift in the process average.
3. Six or more points in a row steadily increasing or decreasing. This signals a trend, such as tool wear.
4. Two out of three points near the same control limit. An early warning of a possible shift.
5. Fourteen points alternating up and down. This suggests two alternating sources of variation, such as two shifts or two machines.
6. Points hugging the center line too closely. This can indicate data that has been mixed from different sources or misrecorded.

Not every rule needs to be used on every chart. Using too many increases false alarms. Pick the ones that make sense for your process and stay consistent.

Important reminder: Control limits are not specification limits. A process can be perfectly in control and still produce output that fails customer requirements. Similarly, a process can meet specifications today and still be out of control, meaning a problem is coming. This is why Six Sigma pairs control charts with process capability analysis (Cp, Cpk).

A Simple Example

Imagine a call center tracking average handling time, in minutes, for each day of the month.

• The center line sits at 6.2 minutes.
• The UCL is 7.4, and the LCL is 5.0.
• For most of the month, the points bounce randomly around 6.2. That’s common cause variation, and no action is needed.
• On day 18, a point jumps to 8.1, beyond the UCL.

That point is a special cause signal. The team investigates and discovers that a system update slowed down the CRM tool that day. They fix it, and the process returns to normal.

Without the chart, the team might have blamed agents or launched a retraining program. The chart pointed to the real cause.

Where Control Charts Fit in DMAIC

Control charts show up in more than one phase:

• Measure: to check whether the process is stable before baseline data is trusted
• Analyze: to spot special causes that point to root causes
• Improve: to confirm that changes actually shifted the process (this pairs well with DOE and statistical modelling)
• Control: to monitor the improved process and keep the gains

In the Control phase, the chart becomes part of the control plan. It is typically paired with a response plan that says who acts, and how, when a signal appears. A chart that nobody responds to is just decoration.

Control charts also support wider quality systems. If you want to see how they connect to company-wide quality culture, explore our Total Quality Management and Lean Management courses.

Common Mistakes to Avoid

Even experienced practitioners slip up. Watch for these:

• Using the wrong chart type for the kind of data you have
• Calculating limits from too little data. Aim for at least 20 to 25 subgroups
• Mixing data from different processes on one chart
• Confusing control limits with spec limits
• Reacting to every point instead of real signals
• Recalculating limits too often, which hides real shifts
• Not documenting actions taken when a signal appears

Fixing these basics makes your SPC efforts much more reliable.

Tools for Building Control Charts

You don’t have to draw these by hand. Popular options include:

• Minitab: the industry standard for Six Sigma work
• Excel: fine for simple charts, with templates or add-ins
• JMP and SPSS: for deeper statistical analysis
• Python and R: for automated, large-scale monitoring (our Business Analytics with Python course is a good starting point)

The software matters less than your understanding. Choose a tool you’ll use consistently, and make sure you understand what it’s calculating behind the scenes.

Why Learn SPC and Control Charts Formally?

Self-study will get you far, but structured training helps with the judgment calls: which chart fits, when a signal is real, and how to present findings to leadership. This is exactly why Six Sigma certification puts so much weight on SPC.

At Gravitex Genesys, learners work through control charts, process variation, and statistical process control with practical examples and guided project work, so the knowledge carries over to real jobs, not just exams. Pick the path that matches your level:

• New to Six Sigma? Start with the Lean Six Sigma Green Belt
• Leading projects? Go for the Lean Six Sigma Black Belt
• Want both? Choose the Lean Six Sigma Combo (Green + Black Belt)
• Aiming to lead at enterprise level? See the Master Black Belt

If you’re building a career in quality, operations, or process excellence, this is a skill worth mastering properly.

Key Takeaways

• Control charts in Six Sigma separate normal variation from real problems.
• Statistical process control (SPC) lets you monitor a process in real time instead of inspecting after the fact.
• Choose your chart by data type: continuous or attribute.
• Control chart analysis relies on clear signal rules, not guesswork.
• Control limits describe the process; specification limits describe the customer’s needs.
• A chart only works if someone acts on what it says.

Ready to Master Six Sigma and SPC?

Understanding control charts is a big step toward becoming a confident process improvement professional. Don’t stop at theory. Practice with real data, learn from experts, and get certified.

Explore the Six Sigma training and certification programs at Gravitex Genesys today, or book a free counseling session to find the right course for you. Start learning now and turn process data into real business results. You can also read more guides on our blog.