Making and Interpreting Run Charts

Use this helpful tool to assess process stability and discover patterns

A run chart is a graphical display of data over time. Run charts are used to visually analyze processes according to time or sequential order. They are useful in assessing process stability, discovering patterns in data, and facilitating process diagnosis and appropriate improvement actions.

Creating the run chart

To start a run chart, some type of product, service, or process must be available on which to take measurements for analysis. Measurements must be taken over a reasonable period of time using a calibrated measurement tool that is being monitored with a calibration control chart. A measurement error study must indicate that the measurement process is acceptable during the data collection process. The data must be collected and stored in chronological or sequential order. You may start at any point in the data set and end at any point. To get meaningful results, at least 25 or more samples must be taken over a long enough period of time so all the components of variation are included.

Once the data have been collected in chronological or sequential order, they must be divided into ordered pairs of x and y values.

…

Want to continue?

By logging in you agree to receive communication from Quality Digest. Privacy Policy.

Create a FREE account

Forgot My Password

Comments

Why Not Process Behavior Charts???

In the modern era of laptop and desktop computers with excellent software packages, why not plot the data on a Process Behavior Chart and gain the insight needed to improve the process that creates the results? Run Charts were OK when we had to generate our charts with a slide rule, #2 pencil, and a pad of engineering paper, but these days, with virtually no extra effort on a computer, a Process Behavior Chart can be created. The four primary Western Electric rules are quite adequate for the analysis and everyone can understand them as operational definitions for signals of lack of a reasonable degree of statistical control. No need to ride a rusty old bicycle when Harley-Davidson is available! Vroooooom!!!

What is the difference?

What is the difference between a run chart and a process behavior chart?

Steve, My answer

We could discuss a process behavior chart but the only difference would be
that it has upper and lower natural process limits. Please review the following
comments and see if you understand why it makes sense to talk about run charts.

1.
The famous statistician Yogi Berra once said, "You can see a lot just by
looking". This is different than having the computer tell you that rule 4
was violated on point 25. The first approach builds observational skill,
thinking and understanding of the process. The second approach only requires
that the analyst have a pulse.

2.
Dr. Deming would often asked his students what was the difference between a
point just outside the control limits and one just inside. The answer is
essentially nothing! So the blind application of an algorithmic approach does
not teach people what they need to know to be a skilled SPC practitioner. I
have personally observed situations where the plug and crank method failed and
management concluded that SPC does not work here, even though the problem was
staring the analyst in the face.

3.
The four Western Electric rules are pretty good but Dr. Shewhart originally
used just rule 1 and Dr. Nelson suggested eight rules. In fact, there are
infinitely many ways a process may exhibit instability so I would not be so
confident that four is adequate? Perhaps looking is more reasonable than just
assuming four is adequate.

In
summary, even if you use good SPC software you always need to look at the data
just the way I described in the run chart article. The software can speed the
analysis and help you find things, but if you don't OBSERVE carefully and THINK
it may speed you to the wrong conclusion.

Tests for runs

I missed you including material on Average Run Length (ARL) and Runs Tests for Randomness.

I agree with Mr. Moore's comments up to the point of not doing runs tests. I routinely did them first to see what the unaltered data told me. If I saw any trends or patterns then i would do an Individuals Chart for more information. If not then I would not be "out" anything.