X-bar S Chart (Mean and Standard Deviation)
Monitor a process using larger subgroups, where the standard deviation measures spread more accurately than the range. The X-bar and S chart tracks the subgroup mean and standard deviation together, with control limits calculated from the data.
Create X-bar S Chart →What is an X-bar S Chart?
An X-bar and S chart is a pair of control charts for continuous data collected in larger subgroups. The X-bar chart monitors the process center by plotting subgroup means, and the S chart monitors process spread by plotting the standard deviation of each subgroup.
It is the counterpart to the X-bar R chart, differing in how spread is measured. For small subgroups the range is efficient enough, but as subgroup size grows the range uses only the two extreme values and wastes information. The standard deviation uses every value in the subgroup, making it more efficient and accurate for larger subgroups, conventionally about 10 or more.
As with the X-bar R chart, the two charts are read in order. The S chart's control limits and the X-bar chart's limits both depend on the average standard deviation, so the S chart must be assessed for stability first; only then are the X-bar chart's limits trustworthy.
In plain terms: Same idea as the X-bar R chart, but for bigger batches. When your subgroups have 10 or more measurements, the range (just highest minus lowest) throws away information, so this chart uses the standard deviation instead, which uses every value. Read the S chart first, then the average chart.
The Two Charts
X-bar Chart (Center)
Plots the mean of each subgroup to monitor the process center. Its limits depend on the average standard deviation, so the S chart must be stable first.
S Chart (Spread)
Plots the standard deviation of each subgroup to monitor variation, using every value rather than just the extremes. Read this first.
When to Use
Best for larger subgroups, about 10 or more, where the standard deviation is more efficient than the range.
Key Formulas
Reading the Charts
Read the S chart first. Because the X-bar chart's control limits are calculated from the average standard deviation, an unstable S chart invalidates the X-bar limits. Only once spread is in control do you interpret the center.
On both charts, look for points beyond the limits and for non-random patterns such as runs and trends, using standard control-chart rules. Any signal indicates a special cause that should be identified and addressed to return the process to stability.
Assumptions & Validation
Variable Data in Larger Subgroups
Continuous data collected in subgroups of about 10 or more.
If violated: For small subgroups use the X-bar R chart; for individuals use an XmR chart.
Rational Subgrouping
Subgroups capture only common-cause variation within them.
If violated: Reconsider subgroup formation if within-subgroup variation is inflated.
Adequate Subgroups
Enough subgroups (commonly 20 to 25) to estimate limits reliably.
If violated: Collect more subgroups before fixing the limits.
⚠️ Check assumptions first
As with the X-bar R chart, interpret the spread chart, here the S chart, before the X-bar chart, because the X-bar limits are derived from the average standard deviation. Choosing the X-bar S chart over the X-bar R chart is driven by subgroup size: for larger subgroups the standard deviation captures spread more accurately than the range. And control limits reflect the process, not the specification.
When NOT to Use X-bar S Chart
Small Subgroups
For subgroups of about 2 to 9, the X-bar R chart is simpler and sufficient.
Individual Measurements
When data arrives one value at a time, use an XmR chart.
Attribute Data
For counts or proportions of defects, use a P, NP, C or U chart.
Industry Applications
High-Volume SPC
Monitor processes where larger samples are collected each period.
Precise Variation Tracking
Track spread accurately when subgroups are large enough to warrant the standard deviation.
Process Stability
Confirm statistical control before assessing capability.
Control-Phase Monitoring
Sustain a process by detecting special causes in center or spread.
Frequently Asked Questions
When should I use an X-bar S chart instead of an X-bar R chart?
Use the X-bar S chart when subgroups are large, conventionally about 10 or more measurements. For such subgroups, the range uses only the two extreme values and wastes information, while the standard deviation uses every value and estimates spread more efficiently and accurately. For small subgroups of about 2 to 9, the X-bar R chart is simpler and its use of the range is adequate.
Why does the X-bar S chart use standard deviation instead of range?
The range considers only the largest and smallest values in a subgroup, ignoring the rest. This is acceptable for small subgroups but increasingly wasteful as subgroup size grows. The standard deviation incorporates every measurement in the subgroup, giving a fuller and more efficient estimate of variation. For larger subgroups this efficiency makes the standard deviation, and therefore the S chart, the better choice.
Which chart do I read first, X-bar or S?
Read the S chart first. The control limits of the X-bar chart are calculated from the average standard deviation, so if the S chart is out of control the spread is unstable and the X-bar limits cannot be trusted. Only after confirming the S chart is in control do you interpret the X-bar chart to assess the process center.
How large should subgroups be for an X-bar S chart?
The X-bar S chart is conventionally used for subgroups of about 10 or more. There is no strict cutoff, and some practitioners use the standard deviation approach for subgroups slightly smaller, but the key idea is that larger subgroups justify the standard deviation over the range. Whatever the size, subgroups should be formed by rational subgrouping so within-subgroup variation reflects common causes.
Are control limits the same as specification limits on an X-bar S chart?
No. As with all control charts, the control limits are calculated from the process data and describe the variation the process naturally produces under common-cause conditions. Specification limits come from customer or engineering requirements. A process can be in statistical control yet not meet specifications, so control charts assess stability while capability analysis assesses whether the process meets requirements.
What signals should I look for on the charts?
Look for points falling beyond the control limits and for non-random patterns such as runs of points on one side of the center line, trends, and cycles, using established rules such as the Western Electric or Nelson rules. Any such signal indicates a special cause of variation has entered the process. Identifying and removing that cause returns the process to a stable, predictable state.
Monitor Mean and Spread With Larger Subgroups
Build an X-bar S chart using the standard deviation for spread. Free during Beta.
Create X-bar S Chart →