Exclusive Content & Downloads from ASQ

Don't Use Rank Sum Tests to Analyze Factorial Designs

Summary: [This abstract is based on the authors' abstract.] Previous work has suggested a rank sum test for 8-run, fractional-factorial designs. This article points out two flaws with this recommendation. First, it relies on a backward elimination procedure that is almost guaranteed to include inactive terms when there are no active effects. Second, rank sums will not be independently distributed for the factorial effects, meaning that if there are any true effects they will affect the distribution of the rank sums of all the other terms.

Anyone with a subscription, including Site and Enterprise members, can access this article.

Other Ways to Access content:

Join ASQ

Join ASQ as a Full member. Enjoy all the ASQ member benefits including access to many online articles.

  • Topics: Engineering
  • Keywords: Design of experiments (DOE), Distributions, Factorial designs, Rank sum test, Screening, Unreplicated fractional experiment
  • Author: Mee, Robert W. ;Lu, Xin
  • Journal: Quality Engineering