- graph of data used to determine when a result should be explored as a special cause
. Developed by Dr. Shewart in the 1920's while he worked at Bell Labs. A control chart have lower control limit and upper control limit and a mean line.
Dr. Shewart developed the control chart to aid in effective exploration and improvement. Often the regular variation (common cause variation
) of a process is taken as something special to be examined which wastes time. In addition, if action is then taken based on this exploration, the effects of tampering often exacerbate problems.
Using control charts the only time special cause analysis is used is when the result was above the upper control limit or below the lower control limit. While this is generally true there are other practices to explore when the data shows odd traits (such as 7 points on one side of the mean).
Types of Control Charts:
- X-chart: used with variable data
- p-chart: used with percentage data (binomial)
- u-chart: used with defect count data
- c-chart: used with count data
- Upper Control Limit - the calculated limit above the mean line. Shewart found in practice it should be placed 3 standard deviations above the mean. It is not a spec or customer requirement, rather it is the voice of the process.
- Lower Control Limit - the calculated limit below the mean line. Shewart found in practice it should be placed 3 standard deviations below the mean (it can be placed at zero when it would otherwise be a negative value and a negative value is not possible.