Okay so back to normal distribution. Here’s what you need to know.

What is it?

You find normal distribution when you take all of your data and create a visual representation of the information. You will illustrate when recurring variations show up in your process. It is actually more helpful when you have a distribution that isn’t normal because then you can say ‘Aha it was the 3 hour traffic jam that affected the process’.  When you hear people talk about the curve, this is what they are referring to.

When do you use it?

This is a tool that is best when used as a continuous probability model with measurements that you don’t have to create. Think about the weight of a cargo shipment or the number of a specific product you receive.

Raw scores and Z scores

Each normal distribution will have a raw score which is made up of two parameters: the mean and the standard deviation. The Z score measures how far you varied from a particular point on your data line.  In real terms it means, if you want to see how many errors occurred on the 5^th then standard deviation shows you that.

Why is it important?

The area under the curve shows the proportion of the curve and which tells you how important this data is to your business. Is the curve is small then you now that the distribution occurs within a relatively small set of circumstances which is easier to control within process. A wider distribution shows you that your process can be interrupted by a variety of factors and may need you to keep a close eye on it.

This should be enough to get you started, but if you are unclear and need some direction give us a call and we can get you started.  Happy analyzing!

Range

The first thing to consider is range. Range will tell you the difference between the most obvious observation and the smallest one. This is important because this is where you identify your outliers (variables that are outside the norm, think of road work on a delivery path or a maternity event as an obvious observation). A large range would be the maternity event; it’s so big there is no way to avoid noticing it. A small range would be a traffic event, it may have impact but the impact will not be evenly distributed and it may or may not impact the final result.

There is a measurement range that is good for a sample size of 2, it’s called the inter-quartile range.  For a bigger sample stick to standard deviation; Standard deviation tells you the average number of times a variation occurs from the mean.

By all means as usual this is not a step by step approach to understanding variation, but it is enough of a foundation to have a conversation with your belt about the metrics and what they mean to your organization and its strategic goals. If you need help starting this conversation, give us a call and we will be happy to get you started.

Nominally Scaled Data

This is the most basic scale and basically tells you whether the information is different or not. This applies to your business in the sense that it tells you the baseline in a yes or no format. Think along the lines of ‘does your customer buy product x’? The answer can only be yes or no.

Ordinal Scaled Data

This data applies to data that can be arranged in a specific order but you cannot distinguish what makes the data different. If you are looking for an answer to why a defect is happening, ordinal data is not going to answer that question.

Interval Scaled Data

This is the sweet spot in terms of data analysis, in this scale the data is able to be arranged in a way that tells you why the defect is happening in specific scenarios. Think along the lines of you need to know why you make more sales on Saturdays. You can measure the sales on Saturdays, the specials you offered on Saturday and how many sales corresponded to the specials offered on Saturday.

Ratio Scale Data

This scale is the most advanced analytic method. When you use this method you have data that has an absolute value and when you get a value of 0 is shows that there is no correlation between the variable and the measurement. For example, you have 10 programmers and programmer A completes 20 lines of code, programmer B completes 15 lines of code.  If programmer C actually completes) lines of code, then you can say that no code was completed on that specific day.

Knowing how to analyze data is a big tool in your Six Sigma tool bag.  Now this is not an exhaustive list, but when you sit down to meet with your belt now you know what you need to ask and what the belts information should be telling you.  When you are ready to get started, let us know and we can help you.

There are different types of data and anytime you measure something you going to need how to interpret it.  There are two main types of data: attributive and variable.

Attributive (Qualitative)

Some people call this the most basic form of data, but for business purposes I don’t accept that. Qualitative data is simple in the fact that it is generally data that can be gathered by asking a yes or no question. For example, ‘Did they buy the new product?’ What is limiting about attributive data is that you really can’t analyze the results in a meaningful way, but it can give you a pretty good place to set your focus.

Variable Data

Variable data is also called quantitative and this is the data that you can measure and analyze. In order to decide if data you have is variable ask yourself these questions:

• Can you classify the data and count the results? (Think number of defects for a particular product line)? If you can this is called discrete data and the limitation of discrete data is that it cannot be broken down into smaller measurements to create additional meaning. It’s a one hit wonder.
• Can the data be measured on a time line with meaningful divisions (Think time, production speed, delivery dates etc…) If you can this is called continuous data and it can be divided further to create additional data.

 As with all of these blogs, this is to get you started and statistical data clearly has more to it than one paragraph. But information is the first step and one you know what type of data you have, you have a better idea of what you need to know. Give us call and we can help you create where you need to go next.

In this blog we have talked about a lot about  the tools used in a change project, but I think that it is time we talk about the information that drives the project-the statistics.  The old adage is that the numbers never lie, but when you work with statistics you know that is a relative statement. The numbers can and do lie, usually it comes down to interpretation. So here are some basic things to remember when looking at the statistics your change project will provide.

1. Understand that statistics are the foundations of 6Sigma, when you don’t understand a tool remember that everyone participating in a 6Sigma project will understand the statistics.
2. The purpose of a statistic is to give you a numerical value for the information collected and analyzed. In other words when you measure something the statistics tell you why the measurement is important.
3. The measurements give you a place to start from, whether that is a place to improve or a place to maintain. (Of course in continuous improvement, there is no place to maintain! Always improve!)
4. Statistics serve as a common denominator, like everything else they can be taken to the extreme but they create a common language and equalize knowledge. Everyone can know and understand the same thing based on the numbers. Statistics can eliminate personality flaws, individual perception and provide your team with a singular focus.
5. The single most important purpose of statistics is the ability to create a new group of problem solvers. Once the group has a common goal they can then began brainstorming for innovative solutions thanks to the data. So numbers=innovation!

As with all things 6Sigma the difficulty level lies within the practitioner. If you want a bunch of charts and reports that can only be interpreted by the consultant, you can find that. If you want to understand how the numbers can help you, give us a call and we will get you started.

graur razvan ionut/freedigitalphoto.net

In this blog we have talked a lot about the different tools we can use in our improvement projects, but it’s my professional opinion that knowing why something works is as good as if not better than knowing that it works. In this blog post we need to discuss how you categorize defects. When you have enough data to cast a critical look over your process, you will find yourself face with the inevitable question of ‘how to I classify the variety of errors’?  Well have no fear; here are the 3 most common classifications:

 Controllable

A controllable error or defect is something specific that you can pinpoint and directly affect with improvement.  For example think of the lowering of the thermostat to improve the operating expenses or pressing the collate button on the copier to reduce the prep time for paperwork.

 Procedural

These are errors that occur during the steps of a process, like a safety checklist or a quality control checklist. The easiest way to decide if an error belongs in this category is to ask yourself ‘ is this a routine step to completing a task? If the answer is yes, then it’s procedural.

Noise

This is the trickiest category of all because it is basically a runoff category.  This category is for things that cannot be fit into a specific setting or procedure.  The category is for miscellaneous and arbitrary errors, think things like the amount of noise children make in school. If it is too big to measure and too hard to assess, it is most likely a noise classification.

The trickiest thing about 6Sigma is knowing which tool to use, to that end knowing where to classify potential areas of improvement is even more necessary. This is a great place to start and once you are ready to start categorizing your errors, we can help. Give us a call and we can you started.

Systems FMEA- This type is applied to products/services at the early concept or design level as each step forms a process which in turn forms a system. System FMEA usually focuses on potential design failures. One good example are the multiple iterations of a product development process, the System FMEA would focus on design failures in the product development process.

Design (DFMEA)-This type works at the pre-production level and it is used to analyze failures prior to production. The goal is to detect how failures affect the system and minimize them. Once caveat regarding this type of FMEA is that the information gained from this type of FMEA must be used in the Process FMEA.

Process (PFMEA)-This type works generally in the early stages of production, where quality planning begins. The failures mode and potential defect sources are given a weighted value and racked and stacked according to your organizational goals. Now as more industries adopt 6Sigma, the thought naturally becomes ‘how does this apply to me’? Whenever a process moves from creation to implementation, you can apply 6Sigma tools. The tools will have to be altered to your specific organizational goals but truthfully and good consultant will adapt the tools according to knowledge and goals anyway.

Equipment FMEA-This type is used to spot the failures in the equipment used in the process or in manufacturing a part of the product. The trick is in knowing your equipment on a level that will allow you to notice errors and failures.

Remember the goal of FMEA is to improve quality and reliability while reducing time and cost.  Before you start I would advise that you think about and identify what you are trying to do and why you are looking at that area. In the meantime if you want to get started right away, give SPC a call and we will get you started.

digitalart/freedigitalphoto.net

This week we will continue our discussion on process mapping, I promise it will not go on forever, but it does have a lot of intricacies. Many people think that process mapping is just putting some shapes on a diagram, but it means much more than that. There are 3 levels of process mapping that are commonly accepted among the 6Sigma crowd.

Level 1 –The Macro Process Map

This is typically how management views the processes of the organization; it’s a big picture, future strategy kind of view. It also creates the ability for management to see how to position the organization or resources in a way that complements the product/service being created. This is a high-level map which generally includes:

• Activities that relate to one major process step
• How the process fits into the big picture
• Little specific detail
• Visualizes only major process steps
• Can be used with only a general understanding of the purpose of the process and its steps.

Level 2-Process Map

This is the worker bee process map, where the people who have specific knowledge of the process come in. This is the map that is used to identify all the major steps a worker takes to complete a process. Within Level 2, there are 4 types of process maps:

• Linear Flow- A straight line from beginning to end.
• Swim Lane-shows you who is responsible for what task.
• SIPOC-a little more complicated. It takes five areas: your suppliers, your inputs, your process, your outputs and your customers.
• Value Stream-a specific map that helps to visualize and understand the metrics for the performance of major steps.

Level 3-Process Flow Diagram

Level 3 is not a must because this is a micro process map. It is where you zero in on a specific area and focus on the steps in the process that are causing whatever challenge you are having. When beginning this level you need to ask the following questions:

• Which steps contributed to the problem?
• Where would the problem most likely have occurred?
• Are there elements to the product/service that lend itself to the problem?

These questions help you find the focus that you decided in your problem statement. For this to work you will have to break each step in the process down, most easily using SIPOC. Remember a Level 3 map should include:

• All process flows
• Any set points
• Any standard or automated procedures
• Inputs and outputs (specify if the are controllable or non-controllable)
• Defects per unit
• Yield and rolled throughput yield
• Value and non value added activities

It’s a lot of information, but mapping a process is a fundamental step in your improvement project. It is absolutely critical that you get it right. For more help or more information, give us a call and we will be happy to get you started.

As I said last week, process mapping is one of my favorite 6Sigma tools and the best thing about it is that anyone can do it. Now as with all things 6Sigma it can be as complicated or as easy as you want it to be. It may seem like process maps have a secret language, but this week’s blog helps you decipher the code.

What is it?

Rectangle illustrates an activity within the process. When activities are described in a rectangle they generally begin with a verb.

When you see a diamond, a decision has to be made. These decisions are generally yes/no or go/stop.

An arrow shows you which way the process is flowing and where it is connected.

A parallelogram shows that this step in the process is a data point.

An ellipse shows the start and ending of a process but some people like ovals or circles. I like circles myself, it really doesn’t matter but if you want to get technical ellipses are the Alpha and omegas of process maps.

Some people swear by MS Project or Visio, but the truth is that MS Excel or MS Word is just as effective for producing process map. The meat of this tool is that you illustrate the steps.  I’ve provided an example of one of the process maps I designed for a client, when you first start mapping a process it’s better to focus on something simpler. This process is just an example of what a finished process map looks like.

What should it include?

At a bare minimum it should show how and where the process starts, who/what influences it (inputs) and the end goal/product.   A more desirable map shows cycle times, value and non-value added tasks and activities, decision points, problems with immediate fix capabilities and process control needs.  But that is not a hard and fast rule; your client will dictate what the map needs to show. As you can see from my process, my client wanted a “no-fluff” chart, a map that only illustrated the tasks that actually took place.

Why use it?

Aside from the clarity that comes with visualization, process mapping is good for:

• Training
• Visualizing improvement points
• Understanding root cause possibilities
• Complementing analytical tools with the data it provides
• Identifying what you will need to make improvements.

What doesn’t it do?

Every tool has its limitations and the process map does not give you a silver bullet. It cannot determine the level of variation, but it can determine if there is variation. It cannot stabilize your process but it can illustrate the best place to start looking for improvements. When you are ready to start giving your processes a deeper look, let SPC help get you started.

How does it work?

Like all things 6Sigma it can be as complicated or as simple as you would like for it to be, having said that there are a few steps that I think you should include in your mapping effort.

1. Define what you need to know.

2. Identify one process at a time and take the process from cradle to grave.

3. In the beginning, stick to linear maps and be sure to define decision points.

4.  Identify if different departments/people participate in the process and define those elements.

What does it look like?

Bottom line –a process map should illustrate your steps and show your organization exactly where you are. When you know where you are, you know where to go.  This is a minimum, but give us a call and we can help you with the specifics.