Six Sigma statistics is an expression derived from one of the branches of statistics called process capability. When it was first established, the term referred to the capability of a manufacturing process to output high proportions within specified parameters. It was assumed manufacturing processes led by the guidelines in the short term translated into long-term production that kept defective products under 3.4 defects per million opportunities (DPMO). The end goal of the program is to raise the effectiveness of all production processes to this level. This would ensure not only that quality goods are produced, but also that customer needs are met.

There are professionals of various knowledge levels in the Six Sigma model; the most common are Black Belts, Green Belts, and Yellow Belts.  Students are advanced as one gains understanding and experiences at the different levels through training. Those who are Green Belts, Black Belts and Master Black Belts are required to have a basic understanding of the two branches statistics upon which the program is based.

One of the branches is descriptive statistics. This branch is used to describe data in numbers and illustrations like graphs and charts. Terms that will commonly be used to summarize data include mode, median, mean, standard deviation and variance. Descriptive surveys depict the performance of the whole section of a process from which the data was obtained. These illustrate the standing in terms of the effectiveness of production processes.

The second branch is that of inferential surveys. Here, descriptive statistics are used to make conclusions or deductions. Descriptive surveys are done in the Analyze and Improve phases. They are arrived at using correlation analysis, regression analysis, hypothesis testing and design of experiments (DOE).

In the majority of cases, it is not possible to analyze entire populations so only a part or sample of the population is analyzed which is called sampling strategy. The findings made from the sample are taken to be representative of the whole population. However, the sample findings may be an accurate representation of the entire population or they may not. They are therefore used with a confidence interval factored in.

Comparisons such as proportions, variance, means and others are started with a hypothetical statement on the population that is being studied. The sample surveys are then considered and studies to verify whether or not with a certain amount of confidence and authority that the theory can be taken to be true or not.

These two methods are applied in independent analysis of the different processes that lead to the production of a product. These include for example the process of procurement of raw materials to be used in production, paying for them, receiving them, storing them, dispatching them to the production line, and then the different steps that are followed to produce a product.

Descriptive statistics for each process are determined after inferential surveys are produced to determine if all the processes are being carried out efficiently or if there are delays and disruptions that led to production of substandard goods or late production that agitates customers.  Statistical analysis is a huge component to the Six Sigma Process and should be considered integral to its success.