Saturday, June 30, 2012

Interpreting a Control Chart

By now, you have a good idea of What a Control Chart is, what it is used for and the types of Control Charts. As a Process Owner, you need to properly interpret the Control Chart because that tells you what actions to take, or not to take. Variation always exists, so your job is to separate variation into its two categories of Common and Special Cause. Then you use your centerlines, control limits, and plots of data to make your interpretation.

A common mistake that inexperienced process owners do is - assigning someone to create lots of Control Charts. While the walls become covered with pretty charts, no one has responsibility to interpret and act on the charts. Acting on just one Control Chart is better than plastering the wall with many charts that go unattended.

Interpreting a Control Chart

Interpreting your control chart is not easy because, your future course of action is going to be based on this interpretation. So, it is imperative that we interpret/understand the control chart in the right way. The following flow chart gives you a decision pathway for taking action on your process.

As you can see - Special Causes require special action, while Common Causes require Lean Six Sigma–type improvement.

Special Causes

Any occurrence or pattern in the data that isn’t just random variation is called a Special Cause. Statisticians have created several signals for detecting Special Causes, the most common of which are outliers, shifts, trends, or cycles. Let’s look at each one of these in detail.
1. Outliers - An outlier is any data point that is above the UCL or below the LCL. Because the control limits are calculated from probability theory, an outlier is highly unlikely in a process with just Common Cause variation.
2. Shifts - The likelihood of a stable process generating nine points in a row on the same side of the centerline is like tossing a coin and getting heads nine times in a row. It’s possible, but highly unlikely. Therefore, a run of nine consecutive points all on the same side of the centerline indicates a shift in the mean. This is a strong indicator that the process has changed and warrants investigation.
3. Trends - A trend is defined as six consecutive points, each higher than the previous point or six consecutive points, each lower than the previous point. This indicates a Special Cause with a gradual effect. Look for a process change that began at approximately the start of the trend.
4. Cycles - Repeating patterns, called cycles, are signaled by 14 consecutive points that alternate up and down. This pattern signals cyclical change in the process that is repetitive and certainly warranting investigation. Possible causes include over-adjustment, shift-to-shift variation, and machine-to-machine variation.

The picture below shows you how you can identify the above special causes from your Control Chart.

These signals (outliers, shifts, trends and cycles) apply to all Control Charts
Common Causes

When no Special Cause signals are present, we deem the process to be in control, not out of control. In other words, the process is stable and not changing and only Common Cause variation is affecting its behavior. The example charts we saw in the previous chapters are examples of control charts that have common cause signals.

Prev: Constructing a Control Chart

Next: Acting on Interpretations

No comments:

Post a Comment

© 2013 by All rights reserved. No part of this blog or its contents may be reproduced or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without prior written permission of the Author.


There was an error in this gadget