Statistical Thinking

The term statistical thinking is used to describe the application of statistical techniques and concepts, and other structured techniques, to a wide variety of situations, including process management and problem solving. As such, it involves the use of data in a process of learning and action.

Statistical thinking is necessary for process improvement, in the sense that the key to process improvement involves the understanding and reduction of variation in the process, and this variation manifests itself in data collected from the process. Process variation can typically be classified as either common cause variation or special cause variation, with each type of variation requiring a different strategy for action. Control charts are used to distinguish special cause variation from common cause variation. More generally, control charts can be used to identify and characterize different sources of variation that act on a process. This involves the understanding and effective use of sampling and subgrouping strategies, which are essentially intelligent ways of gathering data from the process that support process understanding and can thus lead to process improvement.

Statistical thinking is necessary to understand the capability of processes to meet requirements. More advanced statistical techniques for process understanding and improvement include the use of experimental designs, regression analysis, and survival analysis. A graphical tool useful for understanding relationships among variables is a scatterplot. Other tools that support statistical thinking include process flow diagrams, Pareto diagrams, and cause-and-effect diagrams.

The acquisition of the ability to do statistical thinking typically involves training in statistical methods and other methods of process improvement, as well as ongoing practice in the application of these techniques.