From the course: Six Sigma: Black Belt

Unlock the full course today

Join today to access over 22,400 courses taught by industry experts or purchase this course individually.

Two-level fractional factorial experiments

Two-level fractional factorial experiments - Minitab Tutorial

From the course: Six Sigma: Black Belt

Start my 1-month free trial

Two-level fractional factorial experiments

- What if you have many factors, let's say five, and you're unsure of whether every one has a significant impact. How big should your experiment be? A full 2K factorial design for five factors will require two to the power of five, or 32, treatment combinations. With two replicates, that translates to 64 runs. And if each run takes 30 minutes, that's 32 hours of experiments. You'll probably be fired before taking away 32 hours of production time. There must be a better way to do a quick screening to determine which factors are significant to the response and which are not. In fact, there is. It's called a screening experiment using a 2K fractional factorial design. Fractional factorial designs enable you to screen a large number of factors to quickly determine which are the most significant. You can then focus on those few factors in a follow-up experiment to optimize the response using full factorial designs or response surface methods. Fractional factorial designs have fewer runs…

Contents