Factorials & Screening
Study factors together, interpret interactions, and use fractions and follow-up runs to learn efficiently.
Topics¶
An interaction means the effect of one factor depends on another. A main effect can hide that dependence.
Coded sign columns turn factorial effects into orthogonal contrasts. Distinguish an effect from its regression coefficient.
Fewer runs mean aliased effects. Interpret an estimated contrast together with its complete alias set.
A saturated fit has no residual degrees of freedom. Follow-up runs should resolve specific ambiguities in the alias structure.