Use Pre-Process Responses for Analyze Variability to calculate and store the standard deviations of your repeat or replicate responses. Then, you can use Analyze Variability to detect differences in the standard deviations across factor settings. This analysis enables you to identify the factor settings that produce less variable results. Before you can use Pre-Process Responses for Analyze Variability, you must create a 2-level factorial design or define a 2-level factorial design and enter your response data in the worksheet.
You can use the standard deviations that Pre-Process Responses for Analyze Variability stores as response variables to analyze and graph in other DOE tools, such as Contour Plot and Response Optimizer. For more information on analyzing your data, go to Analyze location effects and dispersion effects.
For example, an engineer plans to study how time, reaction temperature, and type of catalyst affect the variability of yield. The engineer creates a 2-level factorial design with 4 replicates. After collecting the data, the engineer uses Pre-Process Responses for Analyze Variability to calculate and store the standard deviations of the replicates at each factor level combination. Then, the engineer can use Analyze Variability to determine which factor settings will yield a more consistent yield.