This page contains a glossary of commonly used terms in Statistical Process Control and Continuous Improvement.
| Name | Definition |
| Area of Opportunity | In the context of process behavior charts, the Area of Opportunity refers to the total number of times an event could have been measured. It represents the potential scope for occurrences within a dataset, which is critical for accurate interpretation of variation. If you are measuring the percentage of students passing a test, the area of opportunity would be the total number of students taking the test. A stable area of opportunity means the total number of students remains consistent between data points, while a varying area of opportunity indicates the total changes between data points, potentially affecting how the data should be analysed. See also Count Data |
| C Chart | A C-chart is a specialised process behaviour chart for count data that monitors the number of defects (nonconformities) in a process. It is used to analyze defect rates when the area of opportunity or sample size is constant across all samples. process limits are calculated based on the total number of defects in each sample, assuming a Poisson distribution for defect occurrences. See also Process Behaviour Chart, Count Data, Natural Process Limits. |
| Causal Model | Your model of which input metrics influence your output metrics, and to what degree. Your organisation's causal model is built through repeated experimentation, and is one of the most important forms of knowledge you can have of your organisation. See also Knowledge, Input Metrics, PDSA Loops |
| Centre Line / Process Average | The central line on an XmR chart, coloured red on an Xmrit XmR chart. The default is for this to be the mean of all the values plotted on the chart, but the median can also be used. See also XmR Chart, Natural Process Limits. |
| Common Cause Variation | Another name for routine variation. See also Routine Variation, Exceptional Variation, Special Cause Variation. |
| Continuous Data / Variable Data | Data measured on a continuous scale, such as time, weight, or temperature. Continuous data provides detailed insights into process behavior. See also Variable Data |
| Continuous Improvement | The ongoing effort to improve products, services, or processes. At Xmrit we emphasise the use of PDSA loops as a path to continuous improvement. See also Kaizen, PDSA Loops |
| Control Chart | Another term for a process behaviour chart. See also Process Behaviour Chart. |
| Controllable Input Metrics | The input metrics which you can control and that reliably influence your desired output metrics. These metrics are key drivers of process performance and are often the focus of optimization efforts. See also Input Metrics, Output Metrics. |
| Count Data / Attribute Data | Data representing counts or attributes, such as defectives or nonconformities. Count data is often used in specialised control charts like C, np, p, and u charts. See also c Chart, np Chart, u Chart, p Chart |
| CUSUM Chart | A chart that monitors cumulative deviations from a target mean to detect shifts. It is highly sensitive to small, sustained changes in process performance. See also EWMA Chart, Process Behaviour Chart. |
| Data Sense | Data Sense is the intuitive ability to understand, interpret, and make decisions based on business metrics. It combines repeated exposure to data, recognition of patterns, and an understanding of causal relationships between metrics. Data Sense allows individuals to identify routine versus exceptional variations, predict trends, and assess the impact of actions or external factors on key business metrics. See also Knowledge, Causal Models |
| Dr. Donald J Wheeler | A pioneer in statistical process control and quality improvement methods, and the primary influence for Xmrit. Dr. Wheeler's work laid the foundation for modern SPC practices. His most popular books are Making Sense of Data and Understanding Variation. His writings can be found at www.spcpress.com See also Walter Andrew Shewhart, William Edwards Deming. |
| EWMA Chart | A chart for tracking the weighted average of process data, sensitive to small shifts. It emphasizes recent data while retaining historical context. See also CUSUM Chart, Process Behaviour Chart. |
| Exceptional variation | Non routine variation caused by specific, identifiable factors. Exceptional variation indicates a need for investigation and corrective action. See also Special Cause Variation, False Alarm, Routine Variation |
| False Alarm / False Positive | When a chart indicates exceptional variation incorrectly. False alarms can lead to unnecessary interventions. See also Exceptional Variation |
| Individual MR Chart (IMR Chart) | Another name for the XmR chart, focusing on individual data points. It combines an X chart and an MR chart. See also XmR Chart, X Chart, MR Chart, Shewhart Chart |
| Input Metrics | Metrics or factors which influence the outputs that you actually want to drive. See also Controllable Input Metrics, Output Metrics. |
| Kaizen | A Japanese term for continuous improvement in all aspects of life and business. It is most commonly associated with Lean Manufacturing at Toyota. It emphasizes incremental, sustainable changes. See also Continuous Improvement. |
| Knowledge | Theories or models used to predict outcomes and improve decision-making. Knowledge is generated through experimentation and data analysis. See also Causal Model, Data Sense, Controllable Input Metrics, Output Metrics |
| Lagged Metrics | Metrics which shift only after a signficant delay following change. Most commonly used in reference to output metrics, which lag changes in input metrics. See also Controllable Input Metrics, Output Metrics |
| Lower Natural Process Limit (LNPL) | The lower natural process limit for an XmR chart, set -3σ from the centre line. See also Upper Natural Process Limit, Natural Process Limits. |
| Moving Range Chart / MR Chart | A chart that tracks variation in moving ranges of data points. It is the second chart that makes up XmR chart, alongside the X chart. See also XmR Chart, X Chart |
| Natural Process Limits / Process Limits | Natural process limits define the range which your process will work when influenced by only routine variation. Defined on an XmR chart as ±3 standard deviations from the process average by default. If a datapoint from your process crosses one of your natural process limits it indicates the presence of exceptional variation, which you should investigate. See also Scaling Factors, Control Chart. |
| Normal Distribution | A bell-shaped curve representing data distribution around the mean. See also Sigma, Standard Deviation. |
| np Chart | An np-chart is a specialised process behaviour chart for count data that monitors the number of defective items in a process. It is used to analyze defect rates when the area of opportunity or sample size is constant across groups. process limits are calculated based on the total count of defective items in each sample, assuming a binomial distribution for defect occurrences. See also C Chart, Count Data. |
| Output Metrics | Metrics representing outcomes or results of a process. These are often the focus of process improvement efforts. See also Input Metrics, Lagged Metrics. |
| p Chart | A p-chart is a specialised process behaviour chart for count data that monitors the proportion of defective items in a process. It is used to analyze defect rates when the area of opportunity or sample size varies between groups. process limits are calculated based on the proportion of defective items in each group, assuming a binomial distribution for defect occurrences. See also np Chart, Count Data. |
| PDSA Loops / PDCA Loops | PDSA Loops (Plan, Do, Study, Act) is a structured approach to trial and error problem solving. It requires you to first Plan what you are going to do, Do the plan, Study the results of the plan, and then Act by either incorporating the learnings into a new PDSA loop or stopping the line of investigation. The original naming was PDCA, with C standing for Check. See also Kaizen, Continuous Improvement. |
| Process Behaviour Chart | The term for all charts identifying routine and exceptional variation in processes. See also Control Chart, XmR Chart, u chart, np chart, p chart, c chart, X̄-R Chart, X̄-S Chart |
| Process Control Worldview | A focus on maintaining process stability and control. See also Process Behaviour Chart, Process Control Worldview. |
| Process Limit Rule | One of the three exceptional variation detection rules from Dr. Donald J Wheeler for XmR charts. Defined as when any point lies outside a process limit. It is the highest priority rule. See also XmR Chart, Dr. Donald J Wheeler, Quartile Limits, Centre Line, Run of 8 Rule, Quartile Rule. |
| Quartile Limits | Limits halfway between the centre line and natural process limits on a chart. These define when the quartile rule will be triggered, one of the three detection rules for exceptional variation used at Xmrit. See also Centre Line, Natural Process Limits. |
| Quartile Rule | One of the three exceptional variation detection rules from Dr. Donald J Wheeler for XmR charts. Defined as when 3 out of 4 points are between a quartile limit and a process limit. It is the second-highest priority rule. See also XmR Chart, Dr. Donald J Wheeler, Quartile Limits, Run of 8 Rule, Process Limit Rule. |
| Rare Event Data | Processes where the data comes in very infrequently making it hard to plot on a process behaviour chart e.g accidents in a factory, outages on a website. See also Chunky Data, Process Behaviour Chart |
| Routine Variation | The predictable variation naturally present in your process all the time, and in all processes to some degree. There is no need to react to routine variation when you see it in your process. See also Common Cause Variation |
| Run of 8 Rule | One of the three exceptional variation detection rules from Dr. Donald J Wheeler for XmR charts. Defined as when 8 consecutive points are on one side of the centre line. It is the lowest priority of the 3 rules. See also XmR Chart, Dr. Donald J Wheeler, Quartile Limits, Centre Line, Process Limit Rule, Quartile Rule. |
| Scaling Factors / Bias Correction Constants / process behaviour chart Constants | Constants used to calculate process limits on process behavior charts. These constants are specific to subgroup sizes. For XmR charts the two main scaling factors are 2.660 for the X chart process limits and 3.268 for the MR chart limit. See also Natural Process Limits, Control Chart. |
| Sensible Defaults | Defaults designed to be practical and widely applicable. See also Scaling Factors. |
| Shewhart Chart | Another name for the XmR chart, focusing on individual data points. It combines an X chart and an MR chart. See also XmR Chart, X Chart, MR Chart, IMR Chart |
| Sigma (σ) / Standard Deviation | Standard deviation, often represented by the Greek letter σ (sigma), is a measure of the dispersion or variability of a data set. It quantifies how much individual data points deviate from the mean (average) of the data. See also Normal Distribution, Process Limits |
| Special Cause Variation | Another term for exceptional variation. See also Exceptional Variation. |
| Spikey Data | Data where large spikes dominate the chart. Spikey data can make the limits on your process behaviour chart unreliable. See also Chunky Data |
| Successive Differences | A method for estimating the standard deviation of a dataset using either the average or median of the moving range of that dataset. The successive differences method is used as it is less influenced by exceptional variation if present, giving a more reliable estimate of the standard deviation. Scaling factors are used in conjunction with the moving range to generate the standard deviation estimate. See also XmR Chart, MR Chart, X Chart, Standard Deviation, Scaling Factors |
| u Chart | A U-chart is a specialised process behaviour chart for count data that monitors the average number of defects per unit in a process. It is used to analyze defect rates when the area of opportunity or sample size varies across samples. process limits are calculated based on the defect rate normalized to the sample size, assuming a Poisson distribution for defect occurrences See also c Chart, p chart, np chart |
| Upper Natural Process Limit (UNPL) | The upper natural process limit for an XmR chart, set + 3σ from the centre line. See also Lower Natural Process Limit, Natural Process Limits. |
| Upper Range Limit (URL) | The name for the natural process limit on an MR Chart See also XmR Chart, MR Chart, Natural Process Limit |
| Walter Andrew Shewhart | The founder of statistical process control methods. See also Dr. Donald J Wheeler, William Edwards Deming. |
| Weekly Business Review (WBR) | A weekly review meeting where all the key controllable input metrics and output metrics for an organisation are reviewed at the same time. Famously Amazon has a well developed WBR. The purpose of the WBR is to build a shared causal model of how your organisation works across a leadership team. See also Causal Model, Controllable Input Metrics, Output Metrics |
| William Edwards Deming | W. Edwards Deming (1900–1993) was an American statistician and business philosopher who revolutionized quality management and operational excellence. His System of Profound Knowledge emphasizes understanding variation, systems thinking, human psychology, and epistemology to drive continuous improvement and long-term success. Deming developed Statistical Process Control (SPC) and 14 Points for Management, and influenced Lean Manufacturing and the Toyota Production System. Deming’s ideas shaped modern quality control and operational practices, driving the industrial transformation of post-war Japan and influencing global business practices. Further information can be found at Commoncog. See also Walter Andrew Shewhart, Dr. Donald J Wheeler. |
| X Bar S Chart / X̄-S Chart | The X̄-S chart is a process behavior chart used when multiple samples can be collected simultaneously, forming subgroups of size ≥2. It calculates natural process limits using the range of the subgroups to monitor process stability. This chart is recommended for subgroup sizes greater than 10, as the standard deviation provides a more precise and stable measure of variability for larger sample sizes. See also X-Bar R Chart, Process Behaviour Chart. |
| X Chart | A chart plotting individual data points in a process. It is the main part of an XmR chart, and is paired with an MR chart. See also XmR Chart, MR Chart |
| X-Bar R Chart / X̄-R Chart | The X̄-R chart is a process behavior chart used when multiple samples can be collected simultaneously, forming subgroups of size ≥2. It calculates natural process limits using the range of the subgroups to monitor process stability. This chart is recommended for subgroup sizes between 2 and 10, as the range is a simple and effective measure of variability for smaller sample sizes. See also X-Bar S Chart, Process Behaviour Chart. |
| XmR Chart | The main process behavior chart for tracking individual and moving range data, and the chart primarily used at Xmrit. It consists of an X Chart and an MR chart. See also X Chart, Moving Range Chart. |