Previous Page

|

Next Page

Main-Effects Plot

In any regular fractional factorial design, uncorrelated estimates of the main effects are available. However, you should keep in mind that main effects are aliased with certain interaction effects. If any of the interactions aliased with a main effect are active, the main-effect estimate will be biased. Usually, if a main effect is part of a significant contrast, the main effect is assumed to be significant. The effects in the alias string can be further explored in subsequent experiments, which are designed to "break the alias chain" by ensuring that effects of interest are estimable and free of other effects.