Invited Comment

Article excerpt

Imrey has written an eloquent and, at times, impassioned argument in favor of statistical certification. I believe that in all respects he has hit the mark, and I would only like to add some personal comments from the perspective of a practicing (M.S.) industrial statistician.

In many respects, Imrey's anecdotes can be summarized by saying that statisticians are too often the Rodney Dangerfields of the scientific world: we don't get any respect. If anything, this understates the plight of statisticians in industry. Despite the fact that modern commercial enterprises are increasingly deluged with data, there is scant recognition of the importance of professional statistical practice. Data collection and analyses that impact not only commercial competitiveness but also the public's health and safety are being done by engineers and scientists using ad hoc methodology with no awareness of the need for statistical methods. Examples include design and analysis for product reliability, safety, and toxicity; environmental impact and workplace risks; public health; and product performance. One of the most spectacular examples is probably the failure to have even one competent statistician in all of NASA or its contractors examine the O-ring data the night before the 1986 Challenger disaster. No manager involved in that decision appeared to have had the slightest clue that statistical methods were available and should have been used for such data analysis. These were smart folks. The result speaks grimly for itself.

Will certification solve all such problems? Certainly not. But surely it must be clear to even its most vigorous opponents that the status quo has failed to convey to the public the role and importance of sound statistical practice. We ought to follow our own admonitions in this regard: look at the data! Other professions--engineers, marketing professionals, human relations and health professionals--have certification and/or accreditation procedures that help establish and assure to the public a body of knowledge and experience necessary for competent professional practice. This not only raises visibility and awareness of the professions but helps provide guidance in judging and choosing among practitioners. Indeed, sometimes this is so important that legal sanctions for competent practice are imposed in the form of licensing. As Imrey has so persuasively argued, our obligation to the public to provide some quality control over our practice is no less important.

I believe that certification would greatly benefit industrial, health, and environmental practitioners. It would establish minimal quality standards that would help employers better assess statistical capabilities. It would help those who gained certification achieve professional recognition formerly available only to their colleagues in Other professions. Indeed, I have known well-qualified statisticians who obtained the ASQC's Quality Engineer certification, for which the statistical component is minimal and inadequate for statistical practice, to help them improve their professional status within their organizations. …