Page 401 - Six Sigma Demystified

Part 3 S i x S i g m a To o l S 381

Statistical Process Control (SPC) Charts

Walter Shewhart, working for AT&T’s Bell Laboratories, developed the concept
of statistical process control (SPC) in the 1920s. Statistical process control uses
data from prior observations of a process to predict how the process will vary
in the future. Shewhart offered this definition of statistical control: “A phenom-
enon is said to be in statistical control when, through the use of past experience,
we can predict how the phenomenon will vary in the future.”
The element of time, which is lacking in enumerative statistical tools (includ-
ing histograms, confidence intervals, and hypothesis tests), is a critical param-
eter for investigating process characteristics because processes are, by definition,
occurring over the course of time. Enumerative tools pool the data from the
past into single estimates of the average and standard deviation—the time ele-
ment is lost! An enumerative analysis of this nature is valid only when we can
assume that all the prior data are from a single distribution.
In contrast, SPC uses the element of time as one of its principal components.
Its control charts can be used to verify that a process is stable, implying that the
data are from a single distribution.

When to Use

Measure Stage

• To baseline the process by quantifying the common-cause level of varia-
tion inherent to the process
• As part of measurement systems analysis to evaluate repeatability and

reproducibility

Analyze Stage

• To differentiate between common and special causes of variation

Improve Stage
• To verify the results of process improvement on the process metric

Control Stage
• To monitor the process to ensure the stability of the revised process and
the continued benefit of the improvement