Creating a Randomized Complete Block Design, as for our first blocking example, is straightforward. Under Create Factorial Design, enter the total number of factors, then click on Designs. Select Full Factorial, then specify the Number of replicates for corner points, and choose Number of blocks. For RCBD's these two choices should match (i.e., we had 3 replications in our example, and chose 3 block as well). Click OK, make any of the usual adjustments to the design (Factors, Randomize Runs, etc) and the design should be loaded in your worksheet. Blocking on fractional factorial designs is relatively easy in Minitab as long as we let Minitab supply the effects to confound with block. As usual for a fractional factorial design, enter the total number of factors you want to analyze in the Number of Factors box in the Create Factorial Design menu. Select Designs and choose the appropriate fraction (full factorial, 1/2 fraction, etc.) you would like to set up. Select the appropriate number of blocks, return to the main dialog box and unselect the Randomized Runs box. Return to the main menu, click OK and your fractional factorial design will be selected for you. Minitab will print the alias structure of the preselected design in its Session window. Note that the design generators for the fractional factorial are listed first and then the block generators are listed. In the worksheet, the runs are grouped by block. When running the analayis, make sure to check Include Blocks in the Model in the Terms dialog box.
If you have to specify a custom block (as in the exercise we discussed in class), you need to select 2-level Factorial (Specify Generators), then choose the number of factors based on the number of runs in the experiment, just as you do when specifying a fractional factorial design (i.e., for an 8-run experiment, select 3 factors, even if an additional factor and a block will be specified). After selecting Full Factorial in the Designs dialog box , select Generators and then enter the block generators in the Block dialog box. In our example, the default half-fraction with two blocks had design generator D=ABC and block generator AB but we needed D=BC as our design generator and AC as our block generator. In that case, we simply enter D=BC in the design generator box and and AC in the block dialog box.
After running your experiment, analysis proceeds in the usual fashion. Note that Minitab doesn't treat the block like an effect so that, for instance, the effects plot has d-1 fewer effects than usual (where d is the number of blocks).