Interpreting x bar R Charts: Patterns
Control chart patterns and their likely causes
Bell Laboratories (reported in Grant and Leavenworth, 1988) identified five typical control chart patterns. Listed below each pattern are representative causes. For each pattern, consider causes with origins in methods, materials, environment, machines, and workforce.
Recurring cycles on x bar chart
Temperature or other recurring changes, such as light or blowers turning on periodically.

Worker fatigue. For example, the accuracy of measurements may decline toward the end of a shift.

Different measuring or testing instruments used in order can produce a regular pattern of differences.

Regular rotation of machines or operators; shift changes.

Merging different streams of processes that themselves have cyclic variation may amplify smaller individual variations. Cyclic variation in feedstock.

 

Recurring cycles on R chart
Regular maintenance, such as changing worn tools may produce patterns in the observed variability.

Worker fatigue. A cyclic pattern of worker fatigue may be mirrored by patterns of increased variability.

Loose set screws or worn parts may produce cyclic variation.

Trends on x bar chart
Accumulation of waste products. For example, washing solutions may gradually lose their effectiveness.

Worker fatigue. Units produced per hour may decline.

Gradual deterioration of equipment, such as wear.

Gradual deterioration of environmental conditions.

Changes in the level of personnel training.

Startup conditions, such as a warm-up period.

 

Trends on R chart
Worker fatigue.

Changes in sub-processes.

Gradual change in materials such as feedstock.

Changes in the level of personnel training.

 

 

 

 

Jumps in level on x bar chart
Change in proportions of materials.

Change in worker or machine. New worker or new machine.

Change in methods.

Change in inspection device or method.

 

Jumps in level on R chart
Change in method.

Change in material.

Change in environment. Change in machine

Change in worker.

High proportion of points near or outside limits on x bar chart
Overcontrol. The operator is making unnecessary changes.

Large systematic differences if material quality.

Large systematic difference in test equipment.

Two different processes are being plotted on the same chart. Example: men's' heights and women's heights.

 

High proportion of points near or outside limits on R chart
Mixture of materials of distinctly different character.

Different people plotting on same chart.

Two different processes are being plotted on the same chart.

Stratification or lack of variability
on x bar chart

Incorrect calculation of control limits.

Incorrect sub grouping. The groups are not
made to allow the variation to show through.

The measuring device lacks needed precision.

 

Stratification or lack of variability
on R chart

The samples contain members of different populations.

The measuring device lacks needed precision.

Process simulations home page