
8 • Root Cause Analysis
variable. In the rust study example, the treatment variable rust inhibitor
was not properly controlled; various concentrations of rust inhibitor were
unknowingly used, and the results of the study were masked by the varia-
tion in the rust inhibitor.
Feynman (1988) points out the need for objectivity in using the scientic
method. He warns experimenters to be careful of preferring one result
more than another and uses the example of dirt falling into an experi-
ment to illustrate this hazard; the experimenter should not just accept the
results that he or she likes and ignore the others. e favorite results may
just be t ...