Measure is the second step of the Six Sigma five step process DMAIC. The objectives of Measure are two fold. First is, using GEMBA, factually understand the existing process. For those that have not heard of GEMBA it means “go see” and in Six Sigma we use the term to imply that you need to get up and go out to the process that you are going to improve and actually “see” what is really happening.  You can NOT do this is a conference room or at your desk alone. The process may in fact be all done at your desk but you need to have other with you looking at what is done to collect the actual way things are done. Second, you then compile that data into a characterization of the current state. Many times you will hear this called the process Baseline.

In measure it is important that you capture exactly what is really happening and do not take for granted that everything is being done exactly to procedure. It usually is not. If more than one person is doing this process check them all as you will find they all are doing it slightly different. Different is not necessarily bad but that difference IS the variation in your process.  So to help me (and you) capture the current state I have a series of questions that if you address them will help insure you complete this step. Here they are:

1. 1. How does it work now?
2. 2. What are the Key metrics for this process? Are they valid?
3. 3. What is the Current Sigma level of the process?
4. 4. What are the Detail steps of the process?
5. 5. What is good about the process you want to keep?
6. 6. What are the problems and their causes?

It is very important that you get all of these answered. Some that stand out are identifying key metrics  and identifying what is good about the process. You need to always look and find out how the process is currently being measured. How does any one know that things are going good or bad with this process? They have to have some “measure” they take to get that idea. Yes sometimes it is a gut feel but you will have to back it up with data so if they say it takes to long; time it and see how long it takes. The second key question that many forget is what works well in the process. This is key because you want to make sure that when you improve the bad you have not made things worse in the good areas. Those you do not want to touch.

To answer the question above it may take several tools and techniques to collect the “facts and data” to answer these questions. We like opinions, they help guide us, but we need data that validates the opinions. So in answering these questions many times you will get opinions but you still need to gather the data to show the opinion is true (or false). So here in Measure, there are several good tool and techniques that can help get you that data.

• Descriptive Statistics –These are generally calculated from a sample of  information (data) off your process. They tend to be in three areas of interest about your data group (distribution). They are location, it’s spread and it’s shape.

• Location or Central Tendencies – These are the mean (average), Median (or middle value) and the Mode (the most frequent value).
• Spread or Dispersion – The most popular are Range (The difference or spread between the highest value and the lowest value) or the Standard Deviation a calculated measure of variation around the Mean or Average.
• Shape – Two things determine shape. One is it Skewed to one side or the other (you can calculate this) and two is it flat or peaked (again you can calculate this [kurtosis])

• Dot Plots – Are just that a plot of every data point on a chart.

• Histograms – These are a pictorial of the data. They are created by grouping the data into what is called bins or cells.

• Run Charts – Similar to the dot chart, the run chart plot the data over time, in time sequence.

• Control Charts – Like the Run Charts, Control Charts plot data over time. Unlike Run Charts they have control limits plotted on the chart as well. Although you can plot individual data points like the one below, a control chart many time plots Summary data over time.

• Gauge R&R (Repeatability & Reproducibility) – These studies are important to understanding how much of the variation that you see is due to the process and how much is do to the measurement system (the way the measurements are taken). Many times this is overlooked but you have to understand that everything varies including the way you measure (collect) your data.
• Pareto Charts – This is a special type of Histogram that arranges the Bins or Cells (categories) into an order of highest frequency to the lowest. This is done so one can see what the major categories are.

• Process Observation Log – This is just a log sheet that you list the process steps in order, how long they take and what the yield is at each step. Sometimes people include a column that identifies the step as value added or Non value added.
• Process Flow Diagramming – This is a  diagram draw to show the sequence of steps from the process observation log. People draw these because it is easier to understand and see what is happening in the diagram then the log list.

• DE & UDE – DE (Desired Effect and Undesired Effect) is just a list of those things that are desirable (things you want to keep) and undesirable (things you do not want). This seems simple, and it is, but we forget to write these things down so that when we get into the heat of things we can not remember them unless they are written down.

Plus here are two more that I talked about in Define: Brainstorming and the 5 Whys.

Well there you have it. A little more understanding of the Measure step of the Six Sigma 5 step DMAIC process.

Peter Bersbach

Six Sigma Master Black Belt

Bersbach Consulting

From Process to Profits

1.520.829.0090

Tags: DMAIC, DMAIC Measure

This entry was posted on Thursday, November 5th, 2009 at 1:59 pm and is filed under Introduction to Six Sigma, Six Sigma Tools. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.