Flexible Tools for Custom Design and Evaluation
DOE++ includes a variety of tools to guide you through the process of creating and modifying your design, as well as for analyzing data from an experiment that doesn't fit any of the predefined design types.
Optimal Design Tool
When none of the pre-defined design types are feasible, DOE++'s Optimal Design tool allows you to build a custom design that is optimized for your situation. With this tool, you can specify a regression model and the software will attempt to optimize the design so it minimizes the uncertainty in the regression coefficients, with the goal of avoiding unnecessary runs in your design.
If there are additional considerations concerning the available factor level combinations (e.g., if you want to exclude certain treatments and include multiple replicates of others), the utility can take these into account as well.
Design Evaluation Feature
All of the design folios now include a built-in feature that can help you evaluate an experiment design before you implement it, or select the best design among several different ones.
For example, you can perform a power study to estimate the design's ability to detect a specified amount of effect. This can help to avoid wasting time and resources on an experiment with insufficient detection power.
You can also evaluate a design by examining its alias structure, comparing its optimality to that of other designs, and calculating other relevant values.
Analyze Response Data Without Building a Design
Two folios enable you to quickly analyze data to explore the regression relationships between the predictors/responses and factors of interest — without having to build a design at all.
- The Multiple Linear Regression Folio is a general statistical tool intended for performing a simple analysis of existing data in order to investigate the effects of predictors on normally distributed responses.
- The new Free Form Folio allows you to analyze data for traditional and reliability DOE by manually entering the factor level combinations and observed response values.