Design expert
| Design expert | |
|---|---|
| Basic data
|
|
| developer | Stat-Ease, Inc. |
| Current version | 12 (July 2019) |
| operating system | Windows , macOS |
| category | Statistics software |
| License | proprietary |
| German speaking | No |
| http://www.statcon.de | |
Design-Expert is a commercial, Windows-based software from the US company Stat-Ease for creating statistical test plans ( English Design of Experiments , DoE). Design-Expert supports the optimization of processes and products with the methods of statistical test planning.
Application of the software
Working with Design-Expert is based on the principles of an experimental cycle . At the beginning of the work with Design-Expert, an experiment type is planned that describes the screening of important influencing variables in their effect. The experimental results of such a plan can be evaluated. Based on the evaluation, optimization plans are designed, implemented and evaluated. With the help of Design-Expert, one or more target values can be optimized using a desirability algorithm . This means that manipulated variables for process parameters can be found for which the target variables are at an optimum.
Design-Expert is not a general statistical program. With Design-Expert, only statistical methods that are directly related to statistical test planning can be calculated. The workflow of the software provides for the steps 1. planning an experiment 2. evaluation 3. renewed and optimized planning and 4. renewed evaluation. The analysis of "historical data" can also be used to evaluate unplanned experiments, but this is less convenient than in general statistical programs.
This disadvantage is compensated for by the fact that the planning and evaluation of statistical tests is extremely fail-safe and easy to carry out. The working steps of the test planning and evaluation are massively wizard-based.
Creatable experimental plans
The following types of test plans can be created and evaluated:
- 2 ^ k plans (fully factorial plans)
- 2 ^ kp plans (factional factorial plans)
- Plackett-Burman plans
- Centrally composed plans (RSM = Response Surface Model = effective surface plans )
- Box Behnken plans
- D and IV optimal plans
- Simplex Lattice Mixtures
- Simplex Centroid Mixtures
- Combined plans
- KCV plans
Additional functions
Other functions of the software include
- an interpretation of the results through commented statistical analyzes and context-sensitive help,
- extensive visualization options for the results (2D and 3D),
- Calculation tool to calculate multiple linear boundary conditions for optimal test plans,
- to maximize many target values simultaneously with an optimization algorithm (desirability),
- the identification and graphical representation of the error propagation ( propagation of error , PoE)
- Evaluation of historical data, d. i.e., records for which no plan has been created in Design Expert.
Alternatives to Design Expert
In addition to Design-Expert, statistical test planning can also be carried out with Minitab , Modde , Fusion QbD , Statgraphics , JMP , STAVEX or Visual-XSel.
literature
- Mark J. Anderson, Patrick J. Whitcomb: DoE Simplified: Practical Tools for Effective Experimentation. Productivity Press, 2007. ISBN 978-1-56327-344-5 (book refers to Design Expert)
- Mark J. Anderson, Patrick J. Whitcomb: RSM Simplified: Optimizing Processes Using Response Surface Methods for Design of Experiments. Productivity Press, 2007. ISBN 978-1-56327-297-4 (book refers to Design Expert)
- Wilhelm Kleppmann: test planning . Optimize products and processes. 7th updated and expanded edition. Hanser, Munich et al. 2011, ISBN 978-3-446-42774-7 ( practical series quality knowledge )