Six Sigma DMAIC Process - Control Phase - Control Chart Selection

Keep these points in mind when selecting a control chart:

Variable data: For individual numerical observations, use an individuals and moving-range (I-MR) chart when appropriate. For rational subgroups, an Xbar chart follows the subgroup means while an R (range) or S (standard-deviation) chart follows within-subgroup spread. The diagram uses Xbar-R for subgroup sizes 2–8 and Xbar-S for 9 or more; this is a selection guideline, not a universal cutoff.

Attribute data: First distinguish nonconforming (defective) units from nonconformities (defects). For units, an np chart shows the count when subgroup size is fixed; a p chart shows the proportion and can accommodate varying subgroup sizes. For defects, a c chart shows counts over equal inspection opportunities; a u chart shows defects per unit and accommodates varying exposure. A p or u chart can also be used with fixed subgroup size or exposure. Check that the sampling and distribution assumptions fit the process.

Open Full-Size Diagram ↗

Six Sigma Control Chart Selection Guidelines

Situation	Example	Statistical Test
Continuous data, one figure at a time	For example: A team is tracking the length of time, measured in days, it takes to process claims. They are randomly selecting one completed claim each day and plotting the time it took to complete that one claim.	Individuals and Moving Range I & MR Chart
Continuous data, averages with small sub-group sizes	For example: A team is tracking cycle time, measured in days, for claims processing. They are randomly selecting five completed claims each day and plotting the average cycle time.	Average and Range X bar & R Chart
Continuous data, averages with large sub-group sizes	For example: A process owner is monitoring call length, measured in minutes and seconds. Each day, the average length for all calls received is plotted. The call center receives over 100 calls per day.	Average and Standard Deviation X bar & S Chart

Open Full-Size Diagram ↗

Six Sigma Control Chart Selection – Continuous Y

Situation	Example	Statistical Test
Discrete data, measuring defects with a constant sample size	For example: A team is tracking the number of incomplete fields on applications. Each day, they pull a random sample of fifty applications and review them for completeness. They plot the total number of fields that are incomplete.	Number of Defects (c Chart)
Discrete data, measuring defects with a variable sample size	For example: Another process owner is tracking the number of fields that are incomplete in applications. Each day her team reviews all incoming applications for completeness and plots the number of incomplete fields per number of applications.	Number of Defects per Unit (u Chart)
Discrete data, measuring defective units with a constant sample size	For example: A team is tracking the number of applications that come in with incomplete information. Each day, they pull a random sample of fifty applications and review them for completeness. They plot the number that are incomplete.	Number Defective (np Chart)
Discrete data, measuring defective units with a variable sample size	For example: A team is tracking the number of applications that come in with incomplete information. Each day, they review all applications for completeness and plot the percent that are incomplete.	Fraction Defective (p Chart)

Open Full-Size Diagram ↗

Six Sigma Control Chart Selection – Discrete Y