Split-Plot Design

Title: One Hard-to-Change Factor
Author: Unlisted
Publication Source: Minitab Inc.

Abstract

This technical support document explains the procedure involved in correctly performing the analysis of results from a split-plot experimental design.

Notes

In practice, I find many practitioners encounter this challenge - namely, that a certain variable is "hard-to-change". Typically it is due to reasons beyond their control (e.g. cost, time, etc.) The implications on the results obtained when just treating this "hard-to-change" variable as a blocking term are discussed in this document. A direct illustration of how to create the template as well as the analysis method using the General Linear Model are shown.