The Design and Analysis of D-optimal Split-Plot Designs Using JMP Version 6 Software

Overview

An experiment is a process or study that results in the collection of data. The results of experiments are not known in advance. Usually, statistical experiments are conducted when researchers can manipulate the conditions of the experiment and control the factors that are relevant to the research objectives. For example, a rental car company might compare the tread wear of four brands of tires, while controlling the type of car, speed, road surface, weather, and driver. However, there are situations in which a condition is difficult to manipulate. For example, it might be difficult to change the temperature of a furnace from 400º C to 350º C and then back to 400º C.