First and foremost, a statistical quality objective is understood at ´Six Sigma´. By definition, Sigma corresponds to the standard deviation of a Gaussian normal distribution. The Sigma level can be determined from the number of errors in a process using tables or statistics programs. With four sigma, 6210 errors occur with one million possible errors. A level of six sigma means less than four errors (3.4 PPM), which corresponds to zero error production. In other words: In a single process step, the error rate must not be higher than 0.00034%.
´Six Sigma´ is also the method of the same name, which is basically a quality management system. Its core element is the description, measurement, analysis, improvement and monitoring (Define, Measure, Analyze, Improve and Control - DMAIC) of business processes by statistical means. The objectives are based on financially important parameters of the company and on customer needs. In this way, measures are to be identified and implemented in order to achieve a higher sigma level. Every higher Sigma is more and more difficult to achieve purely mathematically. Typical methods and tools used are the Kano model, the House of Quality HoQ, the Quality Function Deployment (QFD) as well as the Failure Mode and Effects Analysis (FMEA).