Statistical software is not six sigma

photo credit: Ace_NoOne

Six Sigma is about numbers. Six Sigma produces a flood of data about your process that are critical to your success. If you don’t measure it and understand what you are measuring, you can’t manage it. Six Sigma’s clear strength is a data-driven analysis and decision-making process not someone’s opinion or gut feeling.

The value of statistical analysis cannot be underestimated. Through an analysis of all of that data, you begin to understand your process and develop methodologies to identify and implement the right solutions to improve your process. Statistical evaluation of the data identifies key areas which can have an adverse effect on product quality if not controlled. Once you have identified these key areas you can focus your process improvement efforts.

Given the importance of intelligently handling all of this data, you need to find an efficient and powerful method of crunching the numbers. Naturally, you want to avoid the drudgery of manual calculations and save a whole lot of time by using a statistical software application. You may be tempted to use Excel or another spreadsheet application as a calculator and database to store your statistical process control data. However, you will quickly find out that a basic spreadsheet is too cumbersome to handle the volume and sophistication of the data keeping and analysis you need to perform in a Six Sigma project.

Advanced statistical software such as Minitab, Statgraphics, or Dataplot (free!) are very useful if not essential for gathering, categorizing, evaluating, and analyzing the data collected throughout a Six Sigma project. Both Minitab and Statgraphics are powerful full standalone statistical process control software applications for performing statistical analysis. Both are highly recommended for Six Sigma use as they are tools that can help you utilize one of Six Sigma’s biggest advantages: the ability to make better decisions based upon data.

However in the Six Sigma arena, software is like a crutch. Six Sigma practitioners must realize that numbers don’t represent the total information about an event and statistical software doesn’t solve problems.

I’ve seen practitioners who have mastered Six Sigma statistical software. They know the software inside-out. If you give them some numbers, they spit out a solution in no time while convincing you of the correctness of their solution. These people may actually be dangerous because they can initiate a process change based on software, and produce no gain whatsoever for their company. They can use statistical software to justify anything.

I’ve seen people performing process analysis without even knowing what the process is all about. Statistics is about correlation, while engineering is about causation. Correlation doesn’t mean causation, although causation may lead to stronger correlation. Therefore, Six Sigma practitioners must gain process knowledge before they can interpret a statistical analysis correctly. Besides, statistical software can analyze the data using many techniques and produce an analysis summary that can give a false perception of one’s process expertise.

With a good understanding of the process at hand, practitioners can solve many problems using statistical software. However, if there’s a lack of process knowledge, it’s hard to be sure of what’s going on in the process, or what the statistical analysis is revealing. Process knowledge also expedites statistical analysis, because it allows for Six Sigma experts to know which tool would be effective for performing the desired statistical analysis.