An increasing number of quality control managers are using designed ex
periments to assist in process improvement. This paper presents a step
-by-step description of the design and execution of an experiment cond
ucted by employees of the Athens, Alabama plant of Eaton Controls Divi
sion to improve the performance of an industrial thermostat. The paper
includes a detailed account of how experimental factors were determin
ed and the logic and results of the designers' approach to defining ''
good'' factor levels. In addition, the paper presents the results of t
he actual experiment, problems encountered in data analysis, and analy
tical techniques for handling the censored data that arose during the
experiment.