The zeal for “measurement” and “continuous improvement” expressed through numbers has been gaining strength for some years. I have even seen one comment that “nothing in process development was any use unless it was, or could be, measured”.
I maintain that measurements can’t be made, or make any sense at all, unless they derive first from categories which are specific and concise: in other words, vocabulary (categories). For example, in managing risk we have to decide where there is potential for risk: what I call the “point of failure”, essential for relentless root cause analysis.
In identifying potential points of failure we are obliged to look at design, product development, process, documentation, employee calibre, environment, to name a few. The correspondence between description (category) and reality (condition) must be leak-proof, free from ambiguity and cause no delays in judgement.
I can hear the statisticians protesting about the rejection of numbers analysis, but a number does not say what is, except for counting heads and items; it simply indicates a condition which has had to be defined and clarified elsewhere by vocabulary.
The work in defining categories is well worth while if numbers are to provide true value as measurements for quality system assessments and other critical conditions