Archive for the ‘Six Sigma Tools’ Category

Statistical Engineering

Monday, April 11th, 2011

In the movie “The Graduate,” the new graduate is told by a would-be mentor to remember only one word as he heads out into the world: Plastics. Times have changed. Hal Varian, the chief economist at Google says, ‘‘I keep saying that the sexy job in the next 10 years will be statisticians. And I’m not kidding.’’ Statistical methods are being used by a larger cross-section of people in a wider variety of industries than ever before. There are numerous reasons for this. Nearly everyone has what was once considered to be a supercomputer sitting on their desktop. Powerful statistical software is widely available, including popular packages like Minitab, JMP, SAS and SPSS, and extremely powerful free software. Oracle’s Crystal Ball software makes it possible to create a statistical distribution for any cell in a spreadsheet, making statistical simulation a snap. While becoming more sophisticated, the software is also becoming easier to use. Output is increasingly graphical and easier to explain to laypersons. The number of people trained in Lean Six Sigma methods is growing rapidly. There is an enormous amount of data saved in public and corporate data warehouses. The list goes on and on.

But perhaps the most important reason for the ballooning use of statistics is: it works.

If we take Aristotle’s logic as the historical starting point for rational analysis, and Galileo’s experimental method as the next major leap, then statistical methods might be viewed as the next step in applied analysis. Many problems don’t lend themselves to solution by pure logic nor by carefully planned and controlled experimentation. Most organizations, especially in the commercial sector, must deal with so many problems and such a dynamic external environment that they are forced to make quick decisions despite large uncertainty, then move on to the next problem. Statistical methods help these decision makers evaluate the evidence and make better decisions quickly. The tools and technology described in the first paragraph make this easier than ever before.

This situation is much more akin to engineering than it is to pure science. The approach has been termed “Statistical Engineering.” Authors Roger W. Hoerl and Ron Snee describe Statistical Engineering as follows:

“The statistical engineering discipline [is] the study of how to utilize the principles and techniques of statistical science for benefit of humankind. From an operational perspective we define statistical engineering as the study of how to best utilize statistical concepts, methods, and tools and integrate them with information technology and other relevant sciences to generate improved results. In other words, engineers—statistical or otherwise—do not focus on advancement of the fundamental laws of science but rather how they might be best utilized for practical benefit.

This definition goes beyond applied statistics. Statistical Engineering implies the application of statistics in a systematic framework that utilizes technology to create or improve products, processes and services that improve the lives of people. Disciplines such as Lean Six Sigma, Quality Engineering, Reliability Engineering, and others can be said to do this to some degree, but there are other ways to use Statistical Engineering, some quite unexpected. Billy Beane, general manager of MLB’s Oakland A’s and protagonist of Michael Lewis’s book Moneyball, had a problem: how to win in the Major Leagues with a budget that’s smaller than that of nearly every other team. Conventional wisdom long held that big name, highly athletic hitters and young pitchers with rocket arms were the ticket to success. But Beane and his staff, buoyed by massive amounts of carefully interpreted statistical data, believed that wins could be had by more affordable methods such as hitters with high on-base percentage and pitchers who get lots of ground outs. Given this information and a tight budget, Beane defied tradition and his own scouting department to build winning teams of young affordable players and inexpensive castoff veterans. Author Michael Lewis examines how in 2002 the Oakland Athletics achieved a spectacular winning record while having the smallest player payroll of any major league baseball team. Given the heavily publicized salaries of players for teams like the Boston Red Sox or New York Yankees, baseball insiders and fans assume that the biggest talents deserve and get the biggest salaries. However, argues author Michael Lewis, little-known numbers and statistics matter more.

Statistical Engineering is not limited to applied statistics, theoretical statistics have a place too. In a paper published in the April-June 2011 issue of the journal Quality Engineering author Philip R. Scinto offers this list of Statistical Engineering attributes:

  • Meets high-level needs of an organization
  • Work/study for the greater good
  • Use of statistical concepts and tools
  • Collaborative effort with other sciences
  • Integrated with other sciences
  • Documented protocol
  • Activity continuous with sustainable life
  • Improved results

It isn’t necessary that all items on the list be checked off, but the list is useful in evaluating whether an activity qualifies as Statistical Engineering or if it’s merely another clever use of statistics. The important thing isn’t the label we apply, but the improvement that can be achieved by properly using statistical methods along with science and technology to achieve a challenging goal.

GD Star Rating
loading...

The Problem with Swiss Army Knife Control Charting

Thursday, March 31st, 2011

I’m an advocate of using the I-chart as the default control chart. If I am teaching statistical process control (SPC) and can only teach one chart, the I-chart is always the one that I teach. It’s the only control chart I cover in my Lean Six Sigma Green Belt training. It’s the only chart that I teach in Process Excellence Leadership training. It’s the only chart I use if the data I’m looking at are reasonably close to symmetric (note that I didn’t say “normal”,) unless I have some compelling need for greater sensitivity. I teach that the I-chart is the “Swiss army knife” of control charts.

But I still sometimes use other control charts.

The Problem

Organizations don’t do SPC for the fun of it. They do it because it helps them achieve their goals. Organizations exist to produce things of value for the benefit of customers, investors, and employees. They do this by transforming inputs into outputs of higher value via processes. They can do this better if they minimize variability of outcomes, which can best be accomplished by controlling the sources of variation in the inputs and processes. This is where SPC comes in. SPC is a methodology that uses statistical guidelines to help separate “special cause” and “common cause” variation. If a special cause of variation exists, it signals the need to act. Special cause variation is defined as a change of such a large magnitude that its cause can probably be identified if looked for at once. SPC operationally defines such a change as a measurement result more than 3 standard deviations from the process mean for whatever process metric is being monitored.

A problem might exist if the process generates measurements that are highly skewed, even when it is not being influenced by special causes of variations. Such processes are quite common in the real world. For example, nearly all measurements produced by geometric dimensioning and tolerancing are skewed, as are measurements of time-based phenomena such as those encountered in services industries including the healthcare and hospitality industries. Highly skewed distributions produce a relatively high percentage of results more than 3 standard deviations from the mean even if no special causes exist. In other words, they produce many “false alarms” that will trigger a search for a problem when there is no problem. The false alarms may even lead to tampering, thereby causing a stable process to become unstable.

I-Charts Don’t Solve the Problem

The skewed distribution problem is exacerbated by using I-charts. I-charts are relatively insensitive to moderate departures from normality, and very insensitive if the non-normality still produces a symmetric distribution. But for the data described above, this is not the case. If you use the I-chart for these data you will experience many false alarms. It’s just that simple.

The problem is to determine if a process is or is not being influenced by special causes of variation. A process distribution might appear as skewed because of special cause outliers, or because it naturally produces skewed data. The I-chart treats all data beyond 3 sigma as outliers; it doesn’t help you separate the natural, common cause process outcomes from special cause outcomes. Is the point beyond 3 sigma an outlying chicken, or a common cause egg? I.e., is the process being influenced by special causes, or only common causes? If the process data are naturally skewed you can’t answer this question using an I-chart.

A Simple Solution

The solution that I recommend is to begin your investigation with averages charts, also known as x-bar charts. Averages tend to have distributions that are approximately normal, even if the individual values are skewed. This means that, for a process with a skewed distribution that is not influenced by special causes, averages are much more likely to produce results that stay within 3 standard deviations of the mean than I-charts. It’s the best of both worlds: few false alarms, but still sensitive to special causes. If you have a nice run of subgroup averages without a special cause, plot a histogram of the data and see if the distribution looks skewed or symmetric. If the latter, you can use I-charts with confidence. If the former, stick with averages charts, or find a statistician or Master Black Belt to help you find a more advanced solution.

Stable Does Not Mean Normal

Before ending this article, I’d like to address another pet peeve of mine. I believe that too many teachers of SPC obsess on the need for normality. They confuse normality with the absence of special causes, also known as statistical control. I usually attribute this misunderstanding to a lack of experience with the real world, where normal distributions are so rare as to be virtually non-existent. By insisting on normality we encourage tampering and all of the problems associated with this approach to “process management.”

On the other hand, I am also impatient with people who insist that all non-normality be ignored. These individuals advocate using I-charts in all situations, regardless of the risk of false alarms. This attitude may also be due to a lack of experience. However, I’ve seen SPC lose its credibility when concerned process owners look for special causes over-and-over again without finding them. Like the boy who cried “Wolf!”, out-of-control signals become something to ignore. Eventually so does SPC.

My approach, which favors the I-chart but doesn’t make its use dogma, provides a rational middle ground.

GD Star Rating
loading...

When to Use Your Eyeballs, and When Not To

Thursday, March 24th, 2011

Fig. 1-Large and Small Samples of Normally Distributed Data

One of the exercises I assign to students in my training involves creating two histograms from normally distributed random numbers. The results often look similar to those shown in figure 1. When I ask students to comment on their histograms I usually get comments about the averages, spread, and other statistical properties. However, that misses the point I’m trying to teach.

When we do Six Sigma we usually spend a lot of time mining historical data from databases. Sometimes the sample sizes are large, and sometimes they can be quite small. In fact, even large sample sizes can become small when we slice-and-dice them drilling down with various categories and sub-categories in search of CTQs. Statistical software will often automatically fit a normal curve to histograms created from these data. It’s often tempting to use the fitted curves to make an eyeball judgment about the normality of the data. Sometimes this is a good idea, and sometimes it isn’t. If the sample sizes are small, then the curve may not appear to fit the data very well simply because of small sample variation. Witness the top histogram in figure 1 for an example of a curve fitted to a histogram from a sample size of n=20. The histogram looks like a poor fit, but the p-value of a normality test tells us that the fit is pretty good anyway. So we’re probably safe assuming normality and acting accordingly.

The lower curve is fitted to a sample of n=500 data values. It appears to be a much better fit, and the p-value will back this conclusion. But what if the eyeballed curve fit and the p-value disagree?

histogram-and-probability-plot

Fig. 2-Decent Fit but Lousy P-value

Sometimes the fit of the curve is “close enough,” but the p-value will tell you that the fit is awful. Take a look at figure 2. The histogram suggests that the normal curve fits the data pretty well. There are many practical situations where you could use the normal distribution to make estimates and your estimates would be just fine. These are data on the time it takes to complete technical support calls. If you assume normality and you estimate costs or make a decision about process acceptability, your decisions will be essentially correct. However, the probability plot and AD goodness-of-fit statistic clearly show that the data are not normal and that the lack of fit is particularly poor in the tails (p < 0.005.) A closer examination shows that even in the tail areas the discrepancies are fractions of a percent. For example, the normal distribution estimates that 99.9% of all calls will take less than 35 minutes to complete, while the data show about 99.5%. Chances are these differences are of little or no practical importance.

The point is that in the business world we often need to make decisions and then get on to other, more urgent matters. The normal distribution is a handy device for getting quick estimates that are useful for such decisions. If your sample size is relatively large (say 200 or more) then you can go with the normality assumption if the fitted curve looks reasonably good. On the other hand, if you only have a small amount of data, you can still use the normality assumption if the histogram fit looks lousy, providing the p-value of the goodness-of-fit statistic says the normal curve is okay, i.e., if p > 0.05. The normality assumption is so useful that it’s worth using as a default, even if you bend the rules a bit.

GD Star Rating
loading...

How to Calculate Net Present Value with Excel

Tuesday, March 15th, 2011

GD Star Rating
loading...

Why I Hate Hypothesis Testing

Tuesday, February 8th, 2011

If it were up to me, statistical hypothesis inference testing would be entirely replaced by confidence intervals. Both methods provide exactly the same information, however:

  • Confidence intervals are graphical
  • Confidence intervals are able to compare more than two hypotheses
  • In any real-world situation the null hypothesis is always false. If your sample size is large enough you will always reject the null hypothesis. This often leads to positively silly behavior, such as many government regulations. Newspaper headlines are even sillier.
  • The null hypothesis (literally, the hypothesis that there is no difference) is boring and uninteresting. Hypothesis testing promotes poor science by encouraging researchers to run one experiment and compare the results to this boring alternative. It would generally be better practice to develop and compare several hypotheses with each other.
  • Confidence intervals are less confusing to students, lay persons, and quite frankly most statistics instructors

If you do some research you’ll find quite a body of literature complaining about the hypothesis testing approach. This is my small contribution to that cause.

GD Star Rating
loading...

Has the Process Mean Changed?

Tuesday, February 8th, 2011

Here’s an exercise from Pyzdek Institute Green Belt training. At a pharmaceutical company they have developed an IV drip device that has an advertised drip rate of 5 drops per minute. A sample of 10 “drippers” is taken from the process and tested by counting the number of drips that occur during a 10 minute span. The average for each dripper is found by dividing the total drops by 10. The results are (average drops per minute):

4.9
5.1
4.6
5.0
5.1
4.7
4.3
4.7
4.6
5.0

Use the t test to conduct a test of hypothesis and answer this question at a 95% confidence level: “Is the process producing IV drip devices that average 5 drops per minute?” Also use confidence intervals to answer the same question.

This video shows a way to answer these questions using the QI Macros software.

GD Star Rating
loading...

Six Sigma Project Presentations in a Nutshell

Saturday, August 14th, 2010

I’ve reviewed thousands of improvement projects. I’ve lost count of how many project presentations I’ve attended, either for certification purposes or for presentations to leaders. I’ve come to the conclusion that most Green Belts and Black Belts simultaneously present too much information, and not enough information. If I may speak to Green Belts and Black Belts on behalf of leaders and Master Black Belts everywhere, here’s what I’d like to say. What we’re asking is actually very simple, namely how did you apply the Six Sigma process to pursue a real opportunity? In other words, for your project just walk us through the L1 Six Sigma process shown in the figure, and do so in 45 minutes or less. I actually don’t even care if you use a PowerPoint template, or even if you have any slides whatever. I just want to hear a great Six Sigma success story.

Six Sigma L1 Map

GD Star Rating
loading...

Lean Six Sigma Improvement and Work Design

Monday, June 14th, 2010

This article is an excerpt from a lesson in Pyzdek Institute Lean Six Sigma Black Belt training. Future posts will continue the topic.

In previous lessons you learned how to change a traditional batch-and-queue value stream into a lean value stream. Now we will discuss the design of the actual work that will take place within the processes of the value stream. By going a level deeper we will be able to improve the flow of work within the different processes in the value stream. Specifically, you will learn how to design continuous flow work cells. While the discussion here focuses mainly on manufacturing work cells, the lean principles described apply to any work, including that done in administrative, transaction, or services such as healthcare, retail, and so on.

Selecting Subprojects

The first step is to identify subprojects within the value stream. Subprojects, sometimes called project “loops,” are determined by looking at the future state value stream map and choosing groups of related processes in the value stream for improvement analysis. Each subproject will require a different team with its own set of knowledge, skills, and abilities. However, it is desirable to have at least one member of the Lean Six Sigma team who participates on all of the subproject teams. Figure 1 shows a future state value stream maps with subprojects identified.

Figure 1-Subproject “Loops”

Subproject Loops

Once subprojects are identified, the Lean Six Sigma team must decide which to pursue first, second, and so on. As a general rule it is a good idea to begin at the customer end of the value stream and work backwards. This provides the customer with improved service that they can see and feel quickly. Another criterion is that the pacemaker process should be improved early, since it sets the pace for the rest of the value stream. The “Inside-Out Rule” should be observed: get your own house in order before extending your improvement efforts to include the value streams of outside customers and suppliers. Of course, your decision regarding the starting point should also take into account the likelihood that the subproject will have a big impact on the business and its customers.

Don’t think of the future state value stream map as untouchable. If, as you go through the exercise of selecting and prioritizing subprojects, you see an obvious improvement that’s not on the map, revise the map. Remember, the goal is to improve as much and as quickly as possible.

Once the subprojects have been identified and prioritized, treat each of them as you would any project. You may want to review the modules covering project management in the Define phase at this time. For each project find a sponsor (the value stream owner is a good candidate,)  write a charter, select a team, develop a schedule, identify stakeholders, etc..  By now these things will be second nature to you.

Elements of Work

Figure 2 shows the relationship between value streams, processes, operations, workplaces and procedures in the creation of value. The relationship is hierarchical. To implement Lean all levels of the hierarchy are considered. In previous lessons we discussed ways to change value streams by replacing batch-and-queue push scheduling systems with lean value streams where work is scheduled to maximize flow. Several other lessons focus on ways to improve processes, the next level of the hierarchy. For example, by using process maps to see how work flows through processes or by identifying non-value-added work. In designing work cells we will go deeper than the process level and look at the design of operations, including the layout of workplaces and the standard procedures followed to perform the work in each operation. Such operations are known as standard operations, because the way work is performed follows strict standard procedures.

Figure 2-Value Creation  Hierarchy



Value Creation Hierarchy

Value Creation Hierarchy



Processes are distinct sets of operations nested within a value stream. Process improvement has been the topic of numerous lessons in this course and it requires knowledge of the root causes creating process problems. In the context of designing continuous flow work cells in Lean Six Sigma, we focus primarily on the things in a process that inhibit flow, such as

  • Non-value added process steps on the opportunity map
  • The distance people, materials, or WIP travel between process steps (from the spaghetti chart)
  • Changeover, setup and adjustment time (discussed below)
  • Identify the root causes that are creating quality issues that are responsible for scrap, rework, or problems downstream (discussed in later modules)

In Lean Six Sigma we design work cells that improve the process as well as the specific operations within a cell. We get into “nitty-gritty” details of the work itself, considering how materials are handled and moved, fixtures, workplace layout, movement of various workers, etc. The transfer of work elements  (small units of work) between workers is carefully considered. “Work” is the sum of all of the work elements required to create one complete unit through the entire value stream.

GD Star Rating
loading...

Creating Customer Value or should I say Removing Non Value

Monday, April 12th, 2010

I have seen many companies trying so hard to get their employees to work harder creating more value for their customers. Trying constantly to keep a competitive edge over the competition. And yet when they really look around their employees are working already so hard. In fact, I’d say, people are busy 99% of the time trying to do a good job. So how does a company today meet this challenge, it is in the things the people do. The process! It is not the “people” that creates the value, but the activity (process) they do that creates it. You can actually see this but looking at the “things” (paper or product) in the process evolving into customer needs. If you focus on the “things” in the process and NOT the people you will see that those “things sit there not doing a thing 99% of the time. So to increase value to your customers you need to take the time wasted by the things in the process  just sitting doing nothing and remove it.

How do you do that? Simple, buy looking at the entire process. Look at things people are working on. If you see things that no one is working on then you can bet there is no value being added. Those steps/activities should be eliminated thus removing wasted time from the process. This concept is applied using what is called a Value Added Flow Analysis and I am going to quickly give you the “How To” perform one.

Value Added Flow Analysis

  1. Capture all the steps in the entire activity/process from beginning to end.
    1. To do this you follow one of those “things” (paper or product) from the receiving dock to the customers hands.
    2. Record EVERTHING that happens and how long it takes. I mean everything! Including, for example, the “step” of the thing (order) sitting is a briefcase or notebook as it is transported back to the office to be entered into your system. Or the “step” of the thing (your lunch order at a restaurant) sitting on the note pad as it travels to the kitchen. EVERYTHING! This list will be long both in time and steps.
  2. Next you will take this list and look at each of those step and determine if it is value added or not. So how do you  determine if it is value added? Value added steps can be identified by answering three questions about each step. All three questions have to be answered YES! If any are answered no then they are “non value added steps” and need to be put on the list to be elimination or improvement. Here are the three questions:
  1. Does the thing in the process change? If the “thing” is paper was some information recorded on it? If the “thing” is a product was something added to it?
  2. Does the customer care about the change? In other word are they willing to pay for the change that happened to the thing in the process.
  3. Was it done right the first time? Remember that you, as a customer, do not what to pay for mistake or redo’s and you surely do not want to wait for the error to be corrected. This is of no value to you.
  1. Once you have identified all the value added step then you need to eliminate or significantly improve all of the others. In a simple world you would just eliminate all of the non-value added steps. But our world that is not so easy to do, but I do feel you can eliminate about 75% of them.

Non Value Added Step Eliminating:

How can I be so sure that you can eliminate 75% of these steps; experience. Over the years I find over and over again that you can eliminate about 75% of the non value added steps. Look at one of your processes. When you first developed this “process” it was done a certain way. If lucky that way was written down as a procedure. But as time changed so have customer needs and to meet those needs you have adjusted your process. Over time with all the “adjustments” you now have a process that has several steps that are not needed any more to meet old needs that are no longer there. Another example maybe that the “process” has been handed down from employee to employee (no documentation) and each has done it slightly different. So in time the process has shifted from a originally good one to one that is different during which time the customer needs have changed as well. In either case steps have been left that create no value for your customer and need to be eliminated.

Non Value Added Step Improvement:

OK not everything can be eliminated. Why? Because many time we have more than one customer set of values and we have to prioritize, not eliminate, what we are doing. Be careful you are not micro managing something for you own interest and NOT your customers. A good example of a non value added step that can not be eliminated is Taxes. The “Paying Customer” does not care whether you pay them or not. But to stay in business you have to. Some look at the IRS as another customer (although not a paying one). So in these cases you have to look at ways of completing those steps as quickly and correctly as possible.

Well there you have it. How to create value without something new, but by eliminating waste. That is of value to the customer in that it reduces cost without reducing quality and they receive it sooner than expected. If you like this article I have written several others on my blog http://www.sixsigmatrainingconsulting.com/knowledgebase/ . As always, if you have any questions feel free to contact me.

Bersbach Consulting
Peter Bersbach
Six Sigma Master Black Belt
http://sixsigmatrainingconsulting.com
peter@bersbach.com
1.520.829.0090

GD Star Rating
loading...

The Seven Types of Waste a Summary

Thursday, March 18th, 2010

Bersbach Consulting LLC provides Six Sigma training coaching and support across Arizona, including the Tucson, Phoenix, Scottsdale, and Glendale areas. At this time we would like to thank our friends and clients for their support. If you have landed here looking for our Six Sigma training, coaching or support services in Tucson, then please follow this Six Sigma Training link.

You may have seen a couple of posts I have done on the seven types of waste. I have completed seven articles on all seven types of waste you might find in your organization. Below is a listing and a short description for each of the seven types of waste plus a link to the full article. I believe if you read these articles you will have a new way of looking at your business.


The Seven Types of Waste:


Correction – Corrections are and time you redo, rewrite, rework, repair, or scrap something. This can be as simple as rewriting a grocery list. Say you have a grocery list but you want to rearrange the items on it in the order you will encounter them in the store. Even though it will speed things for you shopping it had to be redone instead of thinking of making the list ordered in the first place. Redoing the list did not add any value to you; it took longer to write it a second time instead of doing it right the first time.


Overproduction – Overproduction is when you make too much of something or you perform too much of a service for some one. Have you ever held a meeting and made copies for that meeting? Most people make a few extra, do you? That is overproduction they will end up in the trash. Or have you every as a question about something in a store and the salesman goes on an on answering your question when all you wanted was the simple answer? That salesman was overproducing


Movement of material or information – This type of waste is when you take any material for information and have to move it from one place to another. You may ship it or carry it your self but that movement does not create any value for the customer in fact it is lost time because it delays your product or service from getting to your customer


Motion of employees – This type of waste is when you or an operator has get up and walk or travel to get something to do their job.  Just like movement of materials and information, motion of the operator does not create value. In fact the “thing” in the process is not changing at all


Waiting – This type of waste is when you, other employees, customer, material, or equipment sits idle waiting. Think about all the waiting rooms there are. As a customer do you want to wait? No but we sometime have come to expect the wait. I have been to doctor’s office where the waiting room is empty or full did not matter but in some I was seen on time and other I have waited over an hour.


Inventory or other resources - This type of waste is not just supplies and materials on shelves but also any recourse your company has that is not being utilized. We normal see inventory as parts and supplies sitting on a shelf like boxes of cereal in the grocery store. But here inventory also include equipment that is standing idle or in storage and employees that have skill that are not being used to their fullest.


Processes - This type of waste is when you are doing more than required by the customer. This is a hard one to understand because sometimes doing more for free has a WOW factor for your customers. That is why it is important to know what is of value and what is not. You see sometime you do sometime more that you think the customer wants and they do not care. That is when it becomes a waste.

If your business is located anywhere in the World including the US, Tucson, Oro Valley , Oracle, Phoenix, Glendale, and Scottsdale, Marana, Green Valley Arizona or beyond and you would like to learn more about our Six Sigma training, coaching and support services please call  Bersbach Consulting LLC at 1-520-829-0090  Now!

GD Star Rating
loading...