One Good Idea

# Flip the Clip

by Kevin L. Johnson

In all the discussion and debate about what Six Sigma means, one thing remains clear: Six Sigma is a broad subject. It is daunting for us trainers to try and produce experts called Black Belts (BBs). We want them to understand lean, statistical process control (SPC), gage repeatability and reproducibility (GR&R), design of experiments (DOE) and multiple regression, but we also want them to become change agents who can break down barriers, work well in a leadership role, save the company millions of dollars and display soft skills.

To make our BBs really effective, we must find a way to link these subjects together into one intuitive skill set. For example, BBs know SPC is based on various statistical distributions, but do they know how control charts relate to hypothesis testing? Do they understand how these distributions relate to Genichi Taguchi's loss function? And what do statistical distributions have to do with the y-hat or s-hat model in a multiple regression derived from a DOE? I am not talking about teaching nonvalue added details. I am talking about making strong intuitive connections.

### There Is No Right Answer

To help your BBs do this, go through your training materials and make a list of every topic, symbol, abbreviation and meaningful phrase you can find (see Figure 1). If you use a small font, everything should fit onto one page. Try to include items such as:

• Abbreviations: SPC, DOE, GR&R, ANOVA (analysis of variance) and COPQ (cost of poor quality).
• Symbols: x-bar, sigma, y-hat, s-hat and chi-square.
• Names of statistics: p-value, F, t and r.
• Topics and phrases: change agent, Juran, fractional factorial and central composite design.

Include anything you would expect the student to recognize and understand. Then find a paper clip. Now you're ready.

Lay the list on a table, and flip the paper clip so it lands on the list. Call out whatever word or phrase it lands on--let's say it's COPQ--and write it on the board. Now flip the paper clip again, call out the second term--let's say it's r--and write it on the board.

Then call on a student to discuss how these two terms are related. The student may discuss how r is used in test equipment correlation and how poor test equipment correlation can lead to incorrect test results, which allow bad product to reach the customer, causing the external failure component of COPQ to increase. Or the student may discuss how variation reduction is correlated to reduced COPQ and can be shown using r.

In another example, let's say the clip first lands on p-value and then on Taguchi loss function. The student may discuss how p-values are used to find significant factors in the s-hat model in the multiple regression analysis of a DOE, and the s-hat model can be used to find process settings to reduce variation, which, according to the Taguchi loss function, will result in less loss to the organization.

There is no one right answer. The goal is for the student to associate the two words or phrases and for the other students to make these connections, too. The class may even have additional comments to add. My list has 180 items. If they are taken two at a time, my students have 16,110 combinations to discuss!

This exercise is especially useful at the end of each day of training. The students, and the instructor for that matter, are usually tired and need to unwind. The instructor also needs to make sure the students can relate what's been discussed to what was discussed the day before.

Taking a few minutes to debrief at the end of the day will help the class unwind in a positive way and strengthen the links that are so important to a BB's success.

KEVIN L. JOHNSON is a Six Sigma Master Black Belt and senior trainer at Sony Magnetic Products of America in Dothan, AL. He earned a bachelor's degree in mathematics from Mary Washington College in Fredericksburg, VA. Johnson is a member of ASQ and is a certified quality engineer, quality manager, quality auditor, reliability engineer and software quality engineer.

Out of 0 Ratings