Locked lesson.

Upgrade

About this lesson

A stable process is one in which only random variation exists. A Lean Six Sigma team must eliminate sources of instability before attempting to improve the normal process performance. The key is undertanding common cause and special cause variation..

Quick reference

Process Stability

A stable process is one in which only random variation within measurable bounds exists.  A Lean Six Sigma team must eliminate sources of instability before attempting to improve the normal process performance.

When to use

The Lean Six Sigma team will determine whether the process is stable during the Analyse phase.  Based on this analysis, the team will select the appropriate improvement strategy.   The team will then need to test the improvement to ensure that the new process is stable.

Instructions

A process is stable if there are no special cause variations present.  Although the process is stable, it will still be subject to common cause variation.  Most improvement strategies for a process are to remove the sources of special cause variation, and if that does not achieve the desired performance, then the process must be fundamentally changed to reduce the common cause variation that is inherent in the process design.

Common Cause Variation

Common cause variation is the normal random variation that is associated with any physical system.  It is always present.  Because it is always present and random in nature but within prescribed bounds of magnitude, it can usually be modeled with a normal distribution.  That means that this variation is predictable and the process controls and management should plan for this level of variation and ensure the process is robust enough to accommodate this variation.  Common cause variation can only be reduced by fundamentally changing the process.

Special Cause Variation

Special cause variation is not random in the same way that common cause is random.  It can be traced to a unique and unpredictable cause that was outside the process.  That unique cause may be random, or it may follow a pattern – the point is that it is outside the process.  Special cause variation can be removed from a process by controlling for the conditions that enabled the unique cause to occur.  Special cause variation leads to process instability because it cannot be modeled and predicted.  Therefore, the occurrence cannot be planned for.  An interesting aspect of special cause variation is that sometimes the special cause can create unexpected and unsustainable good performance.  This often occurs when a process change is introduced and everyone is watching every step of the process closely.  With the best people, the best equipment, and careful oversight occurring, everything goes well.  After the introduction, the normal process performance settles in and the variation is much higher than was experienced during the introduction.  That excellent performance was due to the special cause of the extra oversight.

Hints & tips

  • Make sure you understand whether the variation you see is due to special cause or common cause because the improvement strategy for each is totally different.
  • Some people chase common cause variation as if it were special cause variation.  This inevitably leads to tampering and often drives a stable process into a condition of instability.
Login to download
  • 00:04 Hi, I'm Ray Sheen.
  • 00:06 I want to introduce the concept of process stability.
  • 00:09 Understanding this concept will be a big help in the analyze and
  • 00:13 improve phases of a Lean Six Sigma project.
  • 00:16 Stability starts with understanding of the categories of variation.
  • 00:20 There are fundamentally two kinds, or causes, of variation.
  • 00:24 The first is common cause variation.
  • 00:27 It is the random variation that exists in any physical product or process.
  • 00:32 When the only type of variation that is present is variation from common cause,
  • 00:37 the process is stable and predictable.
  • 00:39 That's because this type of variation can be measured and
  • 00:43 modeled with statistical tools.
  • 00:46 Thanks to this modeling,
  • 00:47 the variation can be predicted to be within defined boundaries.
  • 00:50 This variation is very real, but we know what to expect.
  • 00:55 Designing your process to accommodate at the level of common cause variation will
  • 01:00 result in a very stable process.
  • 01:02 Even though there is a minor common cause variation, you have accounted for
  • 01:06 it in the process design and control.
  • 01:10 The other type of variation is from special causes.
  • 01:13 When special cause variation is present, we say that the process is unstable.
  • 01:18 These are causes from unusual situations not part of the normal process
  • 01:22 environment or process control.
  • 01:25 In fact, they are often caused by the lack of process control.
  • 01:29 Management team or process operator decides to do something in the process
  • 01:33 differently, and then tampers with the process performance.
  • 01:36 This type of variation cannot be modeled with statistical tools,
  • 01:40 because it's not predictable.
  • 01:42 We cannot predict when an unusual event will occur.
  • 01:46 In order for a process to become stable,
  • 01:48 the special cause variation must be identified and then removed.
  • 01:53 Let's look a little closer at common cause variation.
  • 01:57 Common cause variation is always present in a physical process.
  • 02:01 The process designer and process operator should expect it and should allow for
  • 02:06 this level of variation within the process tolerances,
  • 02:09 either input or output variation.
  • 02:12 Ignoring or underestimating it can cause problems.
  • 02:16 Common cause variation is random in nature, and therefore,
  • 02:19 we often model it with a normal distribution, the bell-shaped curve.
  • 02:24 The precise value of the variation could not be predicted for
  • 02:27 any unique instance in the process.
  • 02:30 But we can predict the magnitude or the average amount of variation and
  • 02:33 measure the standard deviation or the variation distribution.
  • 02:38 A key point to remember is that it is always there.
  • 02:41 Although the specific instance of common cause variation is random,
  • 02:44 there are still limits to the variation that are not random.
  • 02:48 A statistical analysis of the random variation will show the upper and
  • 02:52 lower magnitude in the variation,
  • 02:55 which means we can't control common cause variation by taking a special action.
  • 03:01 By special action, I mean trying to accommodate it on a piece by piece or
  • 03:05 item by item basis.
  • 03:07 We can't tweak the process as we chase a common cause variation that is
  • 03:10 slightly higher or slightly lower than the last instance of running the process.
  • 03:15 If we do that, we will inevitably make the system unstable.
  • 03:19 Our tampering with the normal process variation becomes a source of special
  • 03:23 cause variation,
  • 03:24 because we're trying to manage the process differently on every instance.
  • 03:30 So let's now take a deeper look at special cause variation.
  • 03:35 When there is special cause variation occurring, a process is not stable.
  • 03:40 By that, we mean you can't confidently predict the normal amount of variation in
  • 03:44 the process output.
  • 03:46 The normal process controls do not react adequately to special cause variation,
  • 03:51 because it's outside what the process planners had envisioned would occur.
  • 03:55 You may be able to react to the special cause variation, but only by
  • 03:59 taking a special action, such as reworking or scrapping the item that was impacted.
  • 04:05 The point is that the item doesn't just follow normal flow to a successful result
  • 04:10 when a special cause occurs.
  • 04:12 Special cause is not random, and actually, that's a good thing.
  • 04:16 It means that we can usually track down the root cause, and
  • 04:20 it's usually a singular root cause.
  • 04:22 Knowing the root cause, a special control or action can be taken to eliminate
  • 04:27 that root cause, and thus eliminate the effect that has on the process.
  • 04:33 But one caution with that, sometimes special cause creates a special good
  • 04:37 process output that can't be repeated by the normal process control and management.
  • 04:42 This often occurs when everyone is watching closely as we try out a new
  • 04:46 process or introduce a new product.
  • 04:49 There's lots of extra attention and help at each step in the process.
  • 04:53 But when that attention goes away,
  • 04:54 the process cannot sustain that same performance level.
  • 04:58 This is referred to as the Hawthorne Effect.
  • 05:01 So let's summarize the discussion with a head-to-head comparison.
  • 05:06 Common cause variation is the normal variation in a process.
  • 05:10 It's random in occurrence, although with a magnitude within predictable bounds.
  • 05:13 It is inherent in the system design, the equipment, the materials, the process,
  • 05:18 the operator skill and training, the measurement systems and controls.
  • 05:22 Because it is a function of the design, it can be modeled and
  • 05:26 the level of variation can be predicted.
  • 05:29 It cannot be avoided or controlled without fundamentally changing the process design.
  • 05:35 So when it comes time to come up with an improvement to a process,
  • 05:38 if you only have common cause variation, you will need to be recommending
  • 05:43 a significant change to a new process that inherently has less variation.
  • 05:49 We contrast that with special cause variation.
  • 05:51 It's not part of the normal process.
  • 05:54 It is unpredictable in both timing and magnitude.
  • 05:58 There is something that has occurred that is outside the normal control of
  • 06:02 the process.
  • 06:03 Something that was not accounted for in the process design.
  • 06:07 This occurrence cannot be mathematically modeled.
  • 06:10 We can calculate the damage after the fact, but
  • 06:12 we can't predict when it will occur or its magnitude.
  • 06:16 Just a quick sidetrack on predictability,
  • 06:18 some special causes create a pattern in the outcome that is predictable.
  • 06:23 But this is a special cause, because it overrides the normal random variation and
  • 06:28 instead forces the process performance to operate in a prescribed manner.
  • 06:33 However, a good thing is that we can often put controls on possible inputs to
  • 06:38 screen out those special causes.
  • 06:40 And therefore, don't need to reinvest in creating a brand-new process.
  • 06:46 A key goal in the analyze phase will be to determine if the variation in
  • 06:50 the process is due to a special cause or common cause.
  • 06:54 Eliminating the special cause will stabilize the process, and
  • 06:57 allow you to put it under predictable statistical control.

Lesson notes are only available for subscribers.

PMI, PMP, CAPM and PMBOK are registered marks of the Project Management Institute, Inc.

Gift this course