This procedure analyses the factorial design to fit a model. You can also generate effects plots residual plots to assess model adequacy.
1. Choose ProcessMA > DOE > Analyse Factorial Design.
2. In Response, select the column containing the response data.
3. Click OK.
Optional
4. In Terms tab:
· In Maximum order of terms, select the maximum order of terms to include in the model.
· In Available terms, double-click on any terms to include it in the model.
· In Selected terms, double-click on any terms to exclude it from the model.
· Check Include centre points, if you want to include centre points in the analysis.
· Check Include blocks, if you want to include blocking in the analysis.
5. In Results tab:
· Check Display fits and residual, to display the table of fits and residuals.
· Check Plot residuals, to display the residual plot.
· Check Show normal plot of effect, to display a normal probability plot of the effects.
· Check Show Pareto plot of effect, to display a Pareto chart of the effects.
· In Alpha, enter the a-level used to determine the significance of the effects.
Note To select a column of data into a textbox, double-click on any of the column names shown in the list on the left of the dialog box while in the textbox.
You want to investigate how temperature, intensity of light and type of catalyst affect the growth of organisms. You created a full factorial design for three factors and with three replicates. You ran the experiments and recorded the data.
1. Open worksheet ProcessMA > Tools > Data Files > DOE.xls.
2. Choose ProcessMA > DOE > Analyse Factorial Design.
3. In Response, select H – Growth.
4. Click OK.
Interpretation
From p-values in the analysis of variance table, it suggests that the main effects and the 2-way interaction are significant while the 3-way interaction is not.
From the p-values in the estimated effects and coefficients, it suggests that only the two main effects (Temp and Light) and the 2-way interaction (Temp*Light) are significant.
The Normal and Pareto plots of the effects also lets you quickly identify the effects that are significant.