Design of Experiments with Minitab
Training Length - 4 Days

Learn the effective design of experiments from an industry expert.  John Lindland has coordinated, run, and analyzed hundreds of experiments.  Avoiding wasted experiment.  Learn to perform sequential experiments and learn from each one.  Some experiment factors cannot produce good results and must be avoided.  When energy is a factor, there is a chance of interactions with other energy related factors.  Avoid the experiment that tries to explain everything at once.


  • Location, central tendency, normal and non-normal distributions
  • Basic distributions
  • Sample statistics and confidence intervals
  • Basic statistical experiments
  • Factorial Experiments (controlled factor levels)
  • Multiple Linear Regression Experiments (uncontrolled or sorted factor levels)
  • Basic Experiments
  • t-test experiment
  • Paired t-test
  • F-test experiment
  • Passive Statistical Process Control (Xbar-R, P-chart)

Identify Factors

  • Each of the following has its best model to find and select relevant factors
  • Material Processing Model
  • Component Manufacturing Model
  • Assembly Model (component, assembly, and multi-system assembly)
  • Component Design Model
  • Assembly Design Model
  • System Design Model
  • Model Degradation Factors

Prioritize Factors

  • Prioritizing inputs, in-scope, and degradation factors against the intended outputs and unintended outputs.
  • Prioritizing multiple potential response variables.  One experiment can be used to study multiple results.
  • Experiment Purpose, Constraints, and Design
  • Clarify the purpose of the experiment (Variation Reduction-Robustness versus Intended Target)
  • Clarify constraints
  • Controllability of Factors

Managing the Experiment

  • Measurement capability for each factor and response variable,
  • Time to produce conclusions,
  • Cost of test samples,
  • Tooling and equipment availability,
  • Length of test,
  • Importance of conclusions,
  • Availability of people,
  • Support of management/executives)
  • Summarizing the experiment, its priorities, and audience
  • Gaining support, planning the experiment controls, and getting started
  • Setting-up, Running, Analyzing, and Concluding the Experiment

Experiment Models and Strategies

  • Screening, Fractional Factorial Designs, Confounding, and Resolution
  • Sequential experiments, screening, mirror image (increase resolution), center-points (curvature).  All information is used and the important responses define the most important factors and interactions.
  • MS-Excel workbook of factorial designs that can, along with the response variables, be copied into Minitab
  • Analysis of experiments: Workshops from the workbook.  Participants can use the workbook to setup the experiments and follow step-by-stem instructions for loading and analyzing one experiment or sequential experiments.
  • How to use randomization of runs and blocking designs to eliminate external factors that can ruin experiment results.
  • How to avoid experiment settings/combinations that produce useless test samples and results (the 7FM energy oval).
  • Multi-level experiments: Minitab exercise with the experiments, response variables, and step-by-step instructions for performing the analysis and interpreting the results.
  • Center-point designs
  • Three level experiments (qualitative responses have a response surface, qualitative shows the difference between two linear surfaces)
  • Central composite experiments
  • Box-Behnken Designs
  • Path of steepest assent to optimal performance
  • Multiple linear experiments: the factor levels are measured but not controlled.  Can be run using normal production test samples.  Can be run periodically to show changes in the process and continually tune process settings.  Dr. E. Deming, “No process is truly stable.”