Create Factorial Design

Overview

This tool generates 2-level designs full factorial for 2- 7 factors, fractional factorials for 2 - 15 factors and Plackett-Burman designs for 2 - 47 factors. You can label your factors, specify if they are numeric or text and assign values for each factor levels.

2- Level Factorial Design: You can add center points to help you detect curvatures in the fitted data. Center points are added for numeric factors and to each block. If there is a mix of numeric and text factors, pseudo center points are added. To take into consideration other incidental factors such as changes in environment conditions, you can add blocks to the design. For example, if the experiment is to be conducted over several days and the environmental conditions varies, observations obtained in the same day can be in the same block. The number of blocks available depends on the number of factors and number of replicates specified.

Plackett-Burman Design: This is a type of 2-level fractional factorial designs that is commonly used to study main effects. Such designs can be used if 2-way interactions are not significant and you only wish to study main effects. A set of possible number of runs can be chosen based on the number of factors. Similarly, you can add center points and blocks. To add blocks, you need to replicate the runs and each replicate is one block.

The design matrix will be stored in a new worksheet. The cells containing the design matrix are protected to prevent unintended modifications. You can perform the experiement and enter data for each run in the columns after the design matrix for analysis later.

How to

At the Excel Menu (For Excel 2007, go to Add-ins first)

Choose ProcessMA > DOE > Create Factorial Design
In Type of Design, choose to create a 2-level factorial design or Plackett-Burman design
For 2-level factorial fesign
1. In Number of Factors, choose the number of factors for the design
2. In Design, choose from the available designs for the number of factors chosen
3. In Number of Center Points, choose the number of centre points to be added per block of the design
4. In Number of Replicates, choose the number of replications for the corner points
5. In Number of Blocks, choose the number of blocks from the possible choices (based on design and number of replicates)
For Plackett-Burman design
1. In Number of Factors, choose the number of factors for the design
2. In Number of Runs, choose the number of runs for the design
3. In Number of Center Points, choose the number of centre points to be added per block of the design
4. In Number of Replicates, choose the number of replications for the corner points
5. Check Block Design, if you want to block the design by replicates
Check Randomize Runs, if you want to randomise the runs in the design
In Factors table, enter the Type, Name, Low levels and High levels for each factor
Click OK

Example

You want to investigate how temperature, intensity of light and type of catalyst affect the growth of organisms. Since you have sufficient resources to perform 24 runs, you want to generate a full factorial design for three factors, with three replicates.

Choose ProcessMA > DOE > Create Factorial Design
In Type of Design, choose 2-level factorial design
n Number of Factors, choose 3
In Design, choose Full Factorial, 8 Runs, 2^3
In Number of Center Points, choose 0
In Number of Replicates, choose 3
In Number of Blocks, choose 1
In Factors table for Factor 1
1. In Type, choose Numeric
2. In Name, enter Temp
3. In Low, enter 10
4. In High, enter 40
In Factors table for Factor 2
1. In Type, choose Numeric
2. In Name, enter Light
3. In Low, enter 100
4. In High, enter 300
In Factors table for Factor 3
1. In Type, choose Text
2. In Name, enter Catalyst
3. In Low, enter P
4. In High, enter Q
Click OK

Results & Interpretation

You collected data and enter into the Response column