Experimental design and optimization - 2^k factorial designs

9 important questions on Experimental design and optimization - 2^k factorial designs

Simulation versus agricultural experiments:

* all factors are controllable
* random number streams controllable
* experimental conditions constant
* very many factors
* sequential observations

What is sensitivity wrt one factor?

run the simulation at various levels of the factor

What is the OFAT approach?

This is the one-factor-at-a-time approach: change only one factor and make simulation runs at each of the levels of the factor of interest to see how the response reacts to changes in this single factor.
* inefficient
* possible interactions not estimable
  • Higher grades + faster learning
  • Never study anything twice
  • 100% sure, 100% understanding
Discover Study Smart

What is 2^k factorial design?

Choose just two levels for each factor, then make simulation runs at each of the 2

What are design points?

Factor-level combinations

Why shouldn't we choose levels not too close or not too far apart?

Levels not too close: --> difference int output insignificant
Levels not too far apart --> risk of masking nonlinearity, interactions

What is the main effect ej of factor j?

the average change in the response due to moving factor j from its “-“ level to its “+” level while holding all other factors fixed.


What is the two-factor interaction effect ej1j2 of factors j1 and j2?


half the difference between average effect of j1 when j2 at + level and average effect of j1 when j2 at − level.

How to determine the significance of effects?

1) Replicate each scenario n times
2) Estimate each effect n times
3) ompute CI of effect with t distribution with n-1 degrees of freedom
4) If CI does not contain zero --> effect is real

The question on the page originate from the summary of the following study material:

  • A unique study and practice tool
  • Never study anything twice again
  • Get the grades you hope for
  • 100% sure, 100% understanding
Remember faster, study better. Scientifically proven.
Trustpilot Logo