Summa Six Sigma

Summa Six Sigma methodologies refers to the application of statistical methods to problems beyond those normally covered in typical courses dealing with Six Sigma. The software for Six Sigma techniques included fall into the following areas:

  1. Multivariate Capability Analysis - For products whose quality is characterized by more than one measurement, assessment of process capability with respect to each variable separately can be very misleading. A multivariate capability analysis by software for Six Sigma determines the joint probability that all variables will be simultaneously within spec. This requires use of a distribution such as the multivariate normal distribution which accounts for the covariances amongst the variables.

  2. Multivariate Process Control Charts - Control charts based on statistics such as Hotelling's T-Squared are extremely important in monitoring processes with multiple response variables whose joint characteristics affect quality. When combined with a dimensionality reduction procedure such as PCA (principal components analysis) or PLS (partial least squares), such charts can be used to detect changes in a process before they have a serious impact on the final product.

  3. Capability Analysis and Control Charts for Non-Normal Data - Not all measurements are normally distributed. When they are not, use of the normal distribution for the estimation of capability indices and the construction of control charts can be very misleading. STATGRAPHICS Centurion adapts its process capability analyses and control charts to any of 27 continuous distributions.

  4. Control Charts for Autocorrelated Data - When a continuous process is sampled close together in time, control charts must be modified to deal with the autocorrelation. ARIMA control charts capture the dynamics of a process and base signals on the residuals after autocorrelation is removed. Cuscore control charts are configured to detect specific out-of-control patterns in the presence of autocorrelation.  summa_3
  5. Multivariate DOE - Process optimization must take into account all of the responses that characterize a product or service. The use of multivariate desirability functions and optimization methods is critical to overall success.


STATGRAPHICS Centurion software for Six Sigma contains all of the Six Sigma methodologies and procedures necessary to implement Summa Six Sigma. For those interested in learning more about the topics described above, a two-day seminar is offered.