Control Charts

http://www.isixsigma.com/offsite.asp?A=Fr&Url=http://www.skymark.com/resources/tools/control_charts.htm

Control Charts

 

Every process varies. If you write your name ten times, your signatures will all be similar, but no two signatures will be exactly alike. There is an inherent variation, but it varies between predictable limits. If, as you are signing your name, someone bumps your elbow, you get an unusual variation due to what is called a "special cause". If you are cutting diamonds, and someone bumps your elbow, the special cause can be expensive. For many, many processes, it is important to notice special causes of variation as soon as they occur.

There's also "common cause" variation. Consider a baseball pitcher. If he has good control, most of his pitches are going to be where he wants them. There will be some variation, but not too much. If he is "wild", his pitches aren't going where he wants them; there's more variation. There may not be any special causes - no wind, no change in the ball - just more "common cause" variation. The result: more walks are issued, and there are unintended fat pitches out over the plate where batters can hit them. In baseball, control wins ballgames. Likewise, in most processes, reducing common cause variation saves money.

Happily, there are easy-to-use charts which make it easy see both special and common cause variation in a process. They are called control charts, or sometimes Shewhart charts, after their inventor, Walter Shewhart, of Bell Labs. There are many different subspecies of control charts which can be applied to the different types of process data which are typically available.

All control charts have three basic components:

• a centerline, usually the mathematical average of all the samples plotted.
• upper and lower statistical control limits that define the constraints of common cause variations.
• performance data plotted over time.

Things to look for:

The point of making control charts is to look at variation, seeking special causes and tracking common causes. Special causes can be spotted using several tests:

• 1 data point falling outside the control limits
• 6 or more points in a row steadily increasing or decreasing
• 8 or more points in a row on one side of the centerline
• 14 or more points alternating up and down

In those charts that pair two charts together, you will want to look for these anomalies in both charts.

The simplest interpretation of the control chart is to use only the first test listed. The others may indeed be useful (and there are more not listed here), but be mindful that, as you apply more tests, your chances of making Type I errors, i.e. getting false positives, go up significantly.

Types of errors:

Control limits on a control chart are commonly drawn at 3s from the center line because 3-sigma limits are a good balance point between two types of errors:

• Type I or alpha errors occur when a point falls outside the control limits even though no special cause is operating. The result is a witch-hunt for special causes and adjustment of things here and there. The tampering usually distorts a stable process as well as wasting time and energy.
• Type II or beta errors occur when you miss a special cause because the chart isn't sensitive enough to detect it. In this case, you will go along unaware that the problem exists and thus unable to root it out.

All process control is vulnerable to these two types of errors. The reason that 3-sigma control limits balance the risk of error is that, for normally distributed data, data points will fall inside 3-sigma limits 99.7% of the time when a process is in control. This makes the witch hunts infrequent but still makes it likely that unusual causes of variation will be detected.

How should you respond to special cause variation that is picked up by your control chart? Find out. 

http://www.skymark.com/resources/responding_to_variation.asp

Responding to Variation: Special Causes

When a process is being affected by special causes of variation, it is called "unstable", or "out of control". Removing special causes when they are harmful (which is most of the time) or integrating them when they are beneficial (which is rare) is an important part of process improvement.

It is easiest to deal with special causes if they are spotted early and the data used to identify them is timely. Tracking down special causes relies heavily on people's memories of what made that occurrence different from all the others. People may quickly forget any unusual circumstances that may have triggered the unusual variation.

When you spot a special cause:

1. The first thing to do is control any damage or problems with an immediate, short-term fix. Be careful not to view this fix as a permanent solution or the process will never be improved.
2. Once a quick fix is in place, search for the cause. Ask people in the process what was different that time. What was out of the ordinary? It might not have been much – an unexpected emergency, a change in schedules, or new materials. The need for this sort of information is part of the reason for collecting very complete data the first time around, noting details and traceability factors about a sample or recorded event.
3. Once you have discovered the special cause, you can develop a longer-term remedy. Most special causes have a negative impact on the output of the process and need to be removed. Occasionally, a special cause can have a positive impact depending on the nature of the process. If this is the case, finds ways to capture and integrate it into the system.

Avoid these mistakes:

• Changing the process to accommodate the special cause. This usually adds cost and bureaucracy.
• Blaming individuals. Not only does everyone makes mistakes, but also chances are that the problem would have occurred regardless of individuals involved.
• Exhorting workers to simply "do better." People can only do as well as the system allows them to do.

It's important not to stop when you've eliminated special causes of variation. You're only halfway, at that point. The next thing is to reduce common cause variation via systematic process improvement.





If your process is in control, is that good enough? No. You have to start by removing special causes, so that you have a stable process to work with. But then comes the real fun, and often the most substantial benefits: it is time to improve the process, so that even common cause variation is reduced. How to do that? Follow this link.

http://www.skymark.com/resources/responding_to_common_cause_variation.asp

Responding to Variation: Common Causes

Just because a process is stable, or in statistical control, does not mean that its results are satisfactory. A process may be very consistent, day in and day out making items that are nowhere near specification limits. Or, as the Japanese have done so successfully, variation can be systematically reduced, even in stable processes, enabling a gradual tightening of specification limits, and an overall increase in product quality at lower cost.

Improving a stable process is somewhat more difficult than improving an unstable process because, by definition, a stable process has no special causes of variation that jump out at you, asking to be investigated. Instead, you are faced with the task of looking at all data about the process, not simply what made one point different from the others.

Common causes of variation often lie hidden within the system, and are sometimes assumed to be unavoidable. Yet it is very possible, and often very rewarding, to improve processes and reduce common cause variation. Experience had shown that, amongst the people in and around the process, there are enough ideas for improvements to make a significant impact, even on a sound process.

There are many different ways to search for and remove common causes. Probably the most well-known is experimentation, but you can also use stratification. Either of these methods may be helped by disaggregation of data. For more on these, choose one of the following:

Experimentation

Experimenting allows you to test a theory or hunch when you have little or no data available. The best guideline for experimentation with a process is the Plan-Do-Check-Act cycle. The PDCA cycle, described by Walter Shewhart and W. Edwards Deming, is essentially an iteration of the scientific method. The scientific method goes way back…Francis Bacon described it in the 1620's, but its roots reach all the way back to the Greek philosophers.

The PDCA cycle stresses experimentation and observation as the means of discovering truth:

• In the Planning stage, the problem is recognized and analyzed, and possible solutions formulated.
• In the Doing stage, the most likely or effective solution is implemented in a test site.
• The Check is used to compare results of the test solution and the original method to see if there are real improvements.
• Acting involves replacing the old method with the successful solution.

You can then return to the beginning of the cycle to explore other possible problems and strive for new levels of improvement.

The search for common causes is just one of the many arenas in which the PDCA cycle can be used. Most generally, it is used to guide overall process improvement, of which searching for common causes might be but a small part.

The PDCA cycle calls for creative thinking and analytic thinking, both essential to process improvement. Creative or divergent thinking encourages many ideas to be considered and new possibilities to be uncovered. Creativity is an important factor because it can break through paradigms and see beyond the current way of thinking about a process. But creativity must be tempered by analysis or convergent thinking that brings the scattered pieces back together in a workable form.

Stratification

Sometimes experimentation is not necessary, and common causes can be found using stratification of data. Stratifying data is essentially the separation of data into categories: what characteristics are shared or not shared? It often needs to be done iteratively – you stratify at one level, then within one of those categories you stratify again, and so on. If you start at the most general level of the information you have, only the most superficial answer may appear. If you stratify the data at different levels, you may begin to see links. It's like an address on a letter. At the most general level, an address leads you to a country, then to a state, then a ZIP code or city, then to street, then a particular house on the street, and finally to a particular person in the house.

By sorting data into multiple levels of groups with shared characteristics, you can better pinpoint the root cause of a problem. For example, in one 6th grade class, there seems to be a lot of students failing their spelling tests. As you look at the data, you notice that many more students fail the test given on Tuesday than the one given on Friday. When you investigate the Tuesday tests, you notice that most of the failures belong to kids involved with the basketball team. You then learn from the basketball coach that the team holds practice in the evenings on Monday, but immediately after school on Wednesday and Thursday. This means that, most likely, the high number of failures on Tuesdays is due to late practice the night before, leaving the kids less time to study. If you had not stratified the data, becoming more specific at each step, first by days of the week and then by basketball players and non-basketball players, you would never have discovered the common cause in this process.

Stratification can be made easier by using Pareto charts, bar charts, or pie charts, all charts that can display counts of things in different categories. Even the cause & effect diagram could be used to build a tree of branching characteristics, each one being stratified further and further until root causes are reached. In stratification, the characteristic used to separate the data is the "stratification variable." Categories can be composed of a single variable or can combine more than one variable as long as they refer to different characteristics. For example, a category could be "college students who have had their driver's license revoked for underage drinking." This category combine several variables: Are they a college student? Have they had their license revoked? For underage drinking?

Two common mistakes are made when stratifying data. First, it is easy to conclude too much from the stratification. Don't take small differences between category totals too seriously. Look for big differences instead, and try stratifying the data using different stratification variables. Secondly, people tend to jump to the conclusion that an irregularly patterned category is the cause of the problem. The category may provide us a clue as to where to look for the cause rather than being the cause itself.

Disaggregation

Either experimentation or stratification can sometimes be helped by disaggregating the process and viewing its components individually. Sometimes a problem in one part of the process gets covered up by another part of the process. By studying the components separately, a problem that exists in one but is covered up in the whole can rise to the surface.

Disaggregation is not about optimizing each piece at the expense of others. In disaggregation, the parts of the process that are being viewed separately must still be aligned toward the same shared goal and focused on serving the next step in the process. Disaggregation is more about bringing pieces into view rather than actually separating them, or seeing the forest and the trees. Searching for common causes through disaggregation relies heavily on regular meetings between managers of the different parts of the process so that the pieces can be discussed in the context of the whole system.