Important: An existing experimental design (chapter Experimental Data Management) is the prerequisite for any statistical analysis.


The Measuring data manager (compare to chapter Experimental data Management) shows possible measuring data tables as well as Experimental data sheets (see figure), which is the result of an experimental design. When opening an experimental data sheet from the manager, the sheet contains a Statistical analysis button. For the current active experimental data sheet the statistical analysis can also be started from within the main menu (see figure).

Button: Statistical analysis

Two of the four tabs in the window Statistical analysis show the experimental design (D.o.E., see figure) and the resulting Experimental data sheet (see figure). Starting the statistical analysis from the manager allows the user to specifically add or delete measuring points like it is described in the chapter Experimental Data Management.

Figure: Statistical analysis

Figure: Experimental data calculated for the response values power and residence time with the control values screw speed, pressure and feed rate

The actual Response analysis takes place on the third tab in the statistic analysis window (see figure). The user chooses the target variable he wants to take a closer look at as well as the regression model (linear/ quadratic/ cubic). The maximum number of control variables (variables of the model) corresponds with the number of rows in the experimental data sheet. Furthermore, the hierarchy of the model has to be considered, a variable A2, for instance, cannot become part of the regression equation without variable A. The bottom part of the window shows the quality of the regression and because of the graph (bottom right), being two-dimensional, one control variable has to be set.

Figure: Response analysis of the power with the control values screw speed, pressure and feed rate

The fourth tab (see figure) allows an optimization of all target variables at a time. The user enters upper and lower limits for the target variables to be optimized, and the diagram shows those variables in dependence on the control variables. The red-colored part represents the sector where all target variables are within the given limits.

Pointing with the mouse to any part of the diagram the target and control variables of the respective point are shown next to the diagram on the right. By double clicking on the graph the user can create a new process that reflects variables chosen like this.

Figure: Graphic optimization of the response Values Power and Residence Time with the Help of the Control Values Screw Speed, Pressure and Feed Rate.