Lions and Tigers and Variation, oh My!

lean six sigma process controls process management process variation range six-sigma standard deviation statistics
 

In the kids’ movie Madagascar: Escape to Africa, Marty is the zoo zebra suddenly finds himself in a giant herd of zebras and he realizes that he looks just like all the other zebras, and he quickly loses his identity. 

If every individual within a species was identical to every other one, life would be pretty bland. Think about it, every pet dog would look the same, every pet cat would look the same and every person would be identical to every other person. BLAH!  Talk about boring! I guess that is why it is said that diversity is the spice of life!  Variation in our life keeps it interesting, right?

But when is diversity and variation not really a good thing? Let’s think about this. What about jet planes? Do we desire variation between all the Boeing 747’s in the world, or is it better that they are manufactured as identically as possible? What about pilots? Do we desire diversity in their flying methods? Or do we prefer pilots to follow very precise and consistent procedures? We could run through an entire list of things that require consistency, things like medicine, car brakes, and home electricity.

You can see that variation is a good thing in certain aspects of our life and a bad thing in other aspects of our life. When it comes to businesses that sell products, we need both, variety and consistency. Sales and Marketing people strive to create a diverse product portfolio that meets the needs of all the people buying their products. However, on the manufacturing side of things, variation is a bad thing.

Like with Boeing 757’s, customer’s demand product consistency. House siding requires consistent coloring, machined parts require precision, shirt sleeve lengths need to match and light bulbs need to fit the light sockets. You see, in manufacturing, consistency is king. The less product-to-product variation, the better. Tight process controls yield better quality, lower waste and create happy customers.

Variation control is such an important aspect of process management. Reducing and controlling process variation is what the six-sigma methodology is all about. But how do we define and measure it? Well, as an isolated measurement, variation can be measured by range and standard deviation. Range is the simplest form of variation measurement, and it is easily calculated by subtracting the highest point in a data set from the lowest point in the data set. If the highest number is 10 and the lowest number is 4, then the range is 6.

In the discipline of process management, range can a good measurement method if for example it is used in conjunction with process mean in statistical process control (SPC) charting. However, a more useful measurement of process variation is standard deviation. The statistical symbol for standard deviation is the sigma, hence the term six-sigma.

The calculation for standard deviation is a little more complicated than range. In order to determine standard deviation, you need to take these steps:

  1. Calculate the mean or the average of the data set by adding up the data and dividing by number of data points.
  2. Subtract the mean from each individual value.
  3. Square each of those differences.
  4. Determine the average of the squared numbers calculated in step #3 to find the variance. And subtract 1 from the total number of values when finding this average.
  5. Find the square root of the variance.

And that is the formula for standard deviation! But I recommend just using Excel or a statistical software package or an online calculator to find it. Knowing how to calculate a standard deviation is much less important than understanding what it represents and how to analyze it.

The standard deviation is a measure of variation or dispersion. The higher the standard deviation, the greater the variation and that’s a bad thing for a manufacturing process. But the lower the standard deviation the tighter the process variation and that is a really good thing. Ideally, a process operates with a standard deviation of zero because then, every manufactured part would be identical to every other one.  This would provide perfect product quality to every customer and thus making them all very happy.

When it comes to process management and reducing variation, six-sigma methodology is the gold standard. But for a process manager to understand six-sigma, they must understand the normal distribution curve. To learn more about lean and six-sigma, considering enrolling in the "Concepts in Manufacturing" (CML) training courses. The CML courses are effective, flexible and affordable. 

Check out our Training!

Stay connected with news and updates!

If you want some weekly T4T wisdom coming straight to your inbox for your reading pleasure - look no further!  Join our mailing list to receive the latest blogs and updates.
Don't worry, your information will not be shared.

We hate SPAM. We will never sell your information, for any reason.