Adanti, Lee; Walters, Sherry (1988, ASQC) Jeffrey T. Luftig & Associates, Inc., West Bloomfield, MI; Ford Motor Company, Lincoln Park, MI
Typically, statistical methods applications and expertise concentrate on variables (i.e., measurable) characteristics. Diameters, lengths, resistance, time, temperatures, etc. are all measured by assigning a quantitative value thus defining them as variables. Statistically these data, when quantified, will form continuous distributions and hence allow more robust analyses to occur. In contrast, qualitative data is typically evaluated as good or bad and hence defined as attribute data forming discrete distributions. Interestingly enough, final product customer satisfaction most often concentrates on product attributes which in most part are highly subjective in nature.
When confronted with subjective characteristics, there are alternatives to the traditional methods of good or bad evaluations. Attribute data can be quantified. Every characteristic is an attribute until a method of measurement is found which transforms the attribute to a variable. For example, the temperature of a room is an attribute which can be described as either warm or not warm (good or bad). The room temperature stops being an attribute and becomes a variable when a thermometer is used to quantify the attribute.
Various methods that can be used to address subjective data are discussed. A case study is included to demonstrate that once subjective data is quantified and measured it allows statistical evaluations and continuous improvements to evolve. This paper provides an approach to linking customer satisfaction and subjective characteristics which is a mandatory requirement for successful Quality Function Deployment (QFD).
Case study,Statistics,Automobile industry