Bob, et al. --
I've been lurking throughout this conversation. Fascinating diversity of opinion on the subject of the PStat.
Speaking as one of those rogue, unlicensed providers out there -- my academic background is in moral philosophy, not mathematics -- I'm nevertheless called upon in my current profession to provide statistics-informed recommendations about my health plan's quality-improvement programs. We don't do doctoral-level, theoretical statistics filled with more Greek letters than a fraternity party; we make do with techniques like multivariate linear regressions and old-fashioned clinical judgment. Judging by some of the comments I've seen, I think I should start wearing a bib when I fire up SAS or R so that I don't drool on my tie while I'm randomly clicking buttons hoping that "statistics" will result. :)
Whether degreed statisticians like it or not, the line between "stats" and "analysis" is blurry in the business world. If you're a hiring manager, would you rather have to hire two expensive FTEs -- one Ph.D statistician, one data technician -- or one slightly less expensive FTE who will provide a soup-to-nuts solution? (Same holds for solo consultants.) We do have a handful of Ph.D statisticians in my office, but the doctoral credentials were icing on the "jack of all trades" cake. I can provide my own data pulls, scrub my own data, develop content expertise, and apply statistical techniques to arrive at a hopefully sophisticated end product. Professional statisticians rarely, in our experience, have the background in data extraction or the subject-matter expertise to assess the quality of a data set. They just take the numbers and perform their wizardry. Obviously, I'm painting with a broad brush, but I think the larger point holds: What you gain from having a practitioner at the PStat level is offset
in part by skills and experience that the practitioner is less likely to have. The degree to which the "in part" matters ... well, there's the real question the industry will have to grapple with in the coming years.
I'm reminded of my past life as a journalist. Lots of people who advanced to mid- and large-market print publications resisted the ascent of bloggers and the spread of social media. They believed that journalism was a capital-P profession that required strict educational or experiential credentials before you'd be welcomed into the club. The rest of the world said that journalism isn't a profession but rather a practice: Anyone who performs the role of a journalist is a journalist. And although there's not any licensing for journalism in the U.S., the "credential" for a "serious" journalist is probably membership in the Society of Professional Journalists or similar group. I hope statisticians can square the circle on this one and protect sound practice while retaining sufficient humility to recognize that there are many working statisticians out there who are plenty smart and need support even if they don't have a stats degree.
I have no beef with PStat. I just wish that non-degreed statisticians -- those of us who do the work, even if we don't have the sheepskin -- had some mechanism for demonstrating that we're not merely the red-headed stepchildren of the industry.
-------------------------------------------
Jason E. Gillikin, CPHQ
Medical Informatics Consultant
Priority Health
ASA Q&P Webmaster
-------------------------------------------
Original Message:
Sent: 07-30-2013 12:07
From: Robert Riffenburgh
Subject: ASA, Accredited Professional Statistician - what's the point?
Two points.
1) To echo Mark Martin, Richard Erickson, and Janet McDougall, the Royal Statistical Society also has two levels--and I suspect was the first to do so. It is sort of like the Journeyman and Apprentice of two or three centuries ago. That is something the ASA should consider.
2) As I see it, and the reason I became a PStat, is a very long term effort to "professionalize" statistics in the image of non-statisticians. So often statistical analysis is done by people trained in other technical fields. Sometimes it is well done, sometimes not, as is medical care by a non-licenced practitioner. A medical license does not guarantee you'll get good medical care, but it certainly increases the probability. To protect the public from analysis by less-than-professional statistics users, we need a licensing and PStat is a big start in that direction. The point is: look forward 20 years or so and see a requirement for PStat in scientific analysis, legal analysis, statistician hiring ads, etc. That will go far to protect the public from bumblers and charlatans and to protect statistics as a profession from an image of misleading conclusions.
--Bob
-------------------------------------------
Robert Riffenburgh
Naval Medical Center
-------------------------------------------