This hands-on workshop is normally taught in 1 day (or 1.5 days with additional or custom content). Prerequisites: Minitab Essentials and Introduction to Basic Statistics, Factorial Designs, or working knowledge of Minitab and basic classical experimental design methods.

Learn how to handle common DOE scenarios where classic factorial or response surface design and analysis techniques are neither appropriate nor possible due to the nature of the response variable or the data collection process. Develop techniques for experimental situations often encountered in practice such as missing data and hard-to-change factors. Understand how to account for variables (covariates) that may affect the response but cannot be controlled in the experiment. Explore the importance of minimizing process costs while simultaneously optimizing an important process characteristic. Learn how to find and quantify the effect that factors have on the probability of a critical event, such as a defect, occurring.

Topics and Tools Covered Include: ANCOVA, Unbalanced Designs, Split-Plot Designs, Multiple Response Optimization, Binary Logistic Regression.

Optional topics are special additions to standard Minitab training workshops, and extend the course by about 1/2 day. All Minitab training workshops use the latest, current version of Minitab, and authorized Minitab training materials.

#### Analysis of Covariance (ANCOVA)

- define a covariate and understand its rold in a designed experiment
- study the relationship between a covariate and the response variable in a designed experiment
- reduce the error and make factor tests more powerful by including a covariate in a designed experiment
- isolate the effect of the factors in a designed experiment by using a covariate to adjust the response

#### Missing Data in a Designed Experiment

- recognize common scenarios that result in missing data
- identify outliers in a designed experiment
- apply techniques to handle missing data in a designed experiment

Optional: Multi-Vari Charts to visualize results of Factorial Designed Experiments

#### Hard-to-Change Variables

- design an experiment with a hard-to-change factor
- analyze an experiment with a hard-to-change factor
- Cost Optimization
- create a cost column for a process optimization
- analyze data from a central composite design
- determine cost-effective settings using the Multiple Response Optimizer

Optional:

- tree diagrams to visualize nested models
- crossed vs nested factors using tree diagrams
- Multi-Vari charts to visualize relationship between nested and crossed factors, and responses
- discussion of fixed vs random factors, nested vs crossed factors

#### Binary Response

- determine which factors and covariates affect a binary reponse
- interpret the regression coefficients and odds ratios
- display the binary logistic regression prediction equation

#### Plackett-Burman Designs

- use Plackett-Burman designs to efficiently investigate many factors and reduce the impact of confounding
- create and analyze a Plackett-Burman design
- use the Response Optimizer to determine the centre point of a follow-up experiment

#### D-Optimal Designs

- choose an optimal design from a set of candidate points
- check for orthogonality by examining the correlation between factors
- evaluate designs using different types of optimality

#### Practice Exercises

Each workshop participant receives an official Minitab workbook and data files, as well as a certificate for Recertification Units.

See other Minitab Training Courses and comments from previous participants, and some previous clients who have taken Carol Kavanaugh’s workshops!