Previously I wrote a piece attempting to make the distinction between strategy and tactics. The natural follow on question in the context of process improvement is, “Does the implementation of Six Sigma qualify as a strategy or a tactic?”
The answer: It depends.
It may help, when thinking about the distinction between strategy vs. tactic, to conceptualize the two concepts as laying on opposing ends of a continuum. As such, Six Sigma could be considered both a strategy and a tactic, depending on the context of the question.
Execution of a Six Sigma project could be considered a tactic. This may be a specific tactic you use to reach your strategic vision. My black belt project clearly fell into this category as my organization suggested I focus my efforts on reducing the average handle time (AHT) for technical support calls. The strategy was to reduce support costs. The tactic was to run a Six Sigma project aimed at reducing AHT.
On the other hand, you might consider implementation of a Six Sigma program a strategic initiative. Your strategy might include a goal to obtain the highest reputation in the market for quality service. You would likely hire several staff along with a “Director of Quality” and develop policies that require the quality team to be involved in the development of new processes for your service products.
Why does it matter whether Six Sigma is a strategy vs. tactic? You are more likely to successfully accomplish your objectives if you are clear on whether you are using Six Sigma as a tactic vs. strategy. You can avoid the issues associated with ambiguity over whether Six Sigma is core to your business. For example, employees will be less likely to define an initiative as having failed if the initial expectation is to use Six Sigma as a mere tool (tactic) rather than a core strategy. You also avoid the debates over a business’s strategy when the management clearly defines Six Sigma’s role in that organization–even if an organization’s leadership clearly states that Six Sigma is not part of the core strategy.
Once I finished my black belt training, I found that, for better or worse, several peers suddenly considered me an expert in inferrential statistical methods and hypothesis testing. In truth, I was more comfortable using the more commonly used methods in social science (ANOVA, Multiple Regression, etc.) but still felt pretty unsure about the methods I had not yet used like designed experiments (DOE) and auto regressive integrated moving average modeling (ARIMA).
The most common question people seem to ask is, “which statistical test should I use in this case?” I find that many new green and black belts tend to jump to asking what type of statistical test they should use before they have even defined the question they want to answer or developed the hypothesis they wish to test. If you find yourself in this situation, take a step back and write down your very specific question in one sentence. In 99.99% of cases you should be able to state your question in one sentence.
Now back to the original question of which statistical test to use: I don’t have a perfect answer for this question. However, in my black belt training I learned that knowing the types of data (discrete vs. continuous) and how many input or predictor variables you had helped define which statistical tests you could use.
I also found it very useful to find a mentor. But that still leaves two questions: Who makes a good mentor? And, how do I learn from my mentor?
The best mentors for statistical tests and modelling in my experience are econometricians. These are essentially applied statisticians and their training focuses on applying statistical methods to practical business questions. The only trick here is that it could be tough to find one of these rare creatures. If your organization has a marketing or business development analytics group, check there. That’s where you’re most likely to find one.
If you cannot find a mentor, I also recommend a book by Peter Kennedy entitled “A Guide To Econometrics.” Dr. Kennedy’s writing style is far more accessible than most other statisticians and even econometricians. One of the sections in the book is entitled, “The Ten Commandments of Applied Econometrics”. Another section is based on a paper he wrote entitled “Oh No! I Got the Wrong Sign!” Both of these illustrate his practical and approachable style.
The remaining question is how to use your mentor without driving them crazy. My advice here is to do a reasonable amount of research and then present your best guess to him/her on how you should test a particular hypothesis using statistical tests. Most people are far more eager to help enthusiastic newbies that show that they’re really trying. Don’t worry about being wrong. That’s why you’re asking for help.
Working through projects is the best teaching experience. The various tests slowly become second nature. Don’t give up!