SPC control chart types

np, p, c, and µ type control charts are referred to as attribute control charts. These control charts are used when you have "count" type data.

There are two basic types of attributes data—the type of data you have determines the type of control chart to use.

• Yes/No data

• Counting data

To summarize SPC attribute control chart types:

SPC Defective Units​

The SPC Defective Units data source in DataMiner allows you to create control charts that deal with the proportion or fraction of defective units.

A fully-capable process delivers zero (0) defects. Although this may be difficult to achieve, it should always be the goal. Once you resolve the out-of-control point, you can use the quality problem-solving process to begin to eliminate the common causes of defects, for example:

• What are the most common types of defects?

• Why do defects occur?

• What are the root causes of the defects?

Defective Units per Sample (np chart) 

The np control chart is used to evaluate process stability by monitoring the number of defective units in a sample. More information: np chart (Wikipedia.org)

np charts are useful when it's easy to count the number of defective items and the sample size is always the same. Examples of np charts could include:

• The number of defective circuit boards per 100

• Errors per each sample of 50 invoices

np represents the average number of nonconforming units, which can be expressed as n (sample size) times p (the expected proportion of defective units). The data output for a np chart will be the number of defective units or units with symptoms per n units:

The control limits for np charts can be determined using this formula:

The following illustration shows a sample np chart displaying the number of defective units in a specified sample size.

Defective Units per Batch (p chart)

The p control chart is used to monitor the proportion of defective units in a sample, where the sample proportion of defective units is defined as the ratio of the number of defective units to the sample size, n.

The purpose of a p chart is to evaluate process stability when counting the fraction defective. A p chart is used when the sample size varies, for example: the total number of circuit boards or bills delivered varies from one sampling period to the next. More information: p chart (Wikipedia.org)

The data output for a p chart will be the number of defects or symptoms and the sample size. Sample size would be the number of units per each group item, for example: If grouped by job, there will be a sample for each job and the sample size would be the number of units per sample):

When grouping by date/time each sample should be the time increment (for example, month, day, year):

The control limits for this chart type can be determined using the following formula:

The following illustration shows a sample p chart displaying the number of defective units within a sample of batches.

SPC Total Defects

The SPC Total Defects data source in DataMiner allows you to create c and µ control charts that measure defects using fixed and varying sample sizes.

Defects per Unit (c chart)

The c control chart displays the number of defects per unit in samples of a fixed size. ​The purpose of a c chart is to determine the stability of defect counts when the number of units is large compared to the actual number of defects (for example, defects per day, given 2000 units built per day). You can also use c charts to monitor the total number of events occurring in a given unit of time. More information: c chart (Wikipedia.org)

The data output for a c chart will be the number of defects or symptoms per n units:

The calculation for the control limits is:

The following illustration shows a sample c chart displaying defects per unit across a specified sample size.

Defects (µ chart)

The µ control chart displays the number of defects in samples of varying sizes. The purpose of a µ chart is to determine stability of "counted" data (for example, errors per bill, dents in a car door, etc.) when there can be more than one defect per unit and the sample size varies. More information: µ chart (Wikipedia.org)

The µ chart can help you evaluate process stability when there can be more than one defect per unit. Examples might include: the number of defective elements on a circuit board or the number of defects in an invoice or bill.

A µ chart is especially useful when you want to know how many defects there are—not just how many defective items there are. It's one thing to know how many defective circuit boards there are; it's another thing entirely to know how many defects were found within each defective circuit board.

The data output for a µ chart is the number of defects or symptoms and the sample size. The sample size is the number of units per each group item (for example, if grouped by job, there will be a sample for each job and the sample size will be the number of units per sample):



When grouping by date/time, each sample should be the time increment (for example, month, day, year):

SPC chart display attributes

When you use the SPC data sources in DataMiner to create charts, the resulting line charts have the following display characteristics:

• A solid line for the Control Limit

• A dashed line for the Upper Control Limit (UCL)

• A dashed line for the Lower Control Limit (LCL)

For more information about Statistical Process Control and how to use attribute control charts, see the following resources:

• An Introduction to Statistical Process Control (SPC) (Engineering.com)

• Statistical Process Control (Wikipedia.org)