Reason 1: The way statistics is taught in Six Sigma is way outdated. The “tools” are being taught as separate tools rather than showing the students how all of the “classical tools” can be translated and reduced to the simple correlation coefficient (t-test, ANOVA, Regression, Chi-Square). In addition, the tools that organizations really need, i.e. the exploratory, data mining tools are not being taught. There is still the bias that data exploration is a “fishing expedition” and that the only good statistical analysis is based on experimental design. So you end up with frustrated GBs and BBs who have to forcefit their “toolkit” to the problem rather than use the tools to identify the problem. What is “root cause analysis” other than identifying/naming the problem and understanding the mechanisms that create the problem. Common sense has been lost in the narrow-minded fixation on “inferential statistics”.
Reason 2: The way Six Sigma is being deployed. If your goal is to “certify” x GBs and BBs you have a certification process including a “body of knowledge” blessed by the ASQ you get a training program that was developed about 18 years ago and hasn’t changed or adapted since then.
Six Sigma has ossified into a template that is forcefit into each organization. Unless you have someone experienced enough to adapt this methodology to the specific needs of the organization, it can be quite a frustratring experience.
.
