Using Graphical Diagnostics to Deal With Bad Data

ABSTRACT

This article deals with thorny issues that confront every experimenter, i.e., how to handle individual results that do not appear to fit with the rest of the data. It provides graphical tools that make it easy to diagnose what is wrong with response data—damaging outliers and/or a need for transformation. The trick is to maintain a reasonable balance between two types of errors: (1) deleting data that vary only due to common causes, thus introducing bias to the conclusions. (2) not detecting true outliers that occur due to special causes. Such outliers can obscure real effects or lead to false conclusions. Furthermore, an opportunity may be lost to learn about preventable causes for failure or reproducible conditions leading to breakthrough improvements (making discoveries more or less by accident).

Two real life data sets are reviewed. Neither reveals its secrets at first glance. However, with the aid of various diagnostic plots (readily available in off-the-shelf statistical software), it becomes much clearer what needs to be done. Armed with this knowledge, quality engineers will be much more likely to draw the proper conclusions from experiments that produce bad (discrepant) data.

KEYWORDS:

• Box-Cox plot
• design of experiments (DOE)
• diagnostics
• outliers
• transformations