Statistical Consultant
I have acted as an in-house statistical consultant to different departments at UNLV, UN School of Medicine, and various organizations in Las Vegas since 1986.
A Simple Note: Statistical Consulting in a University
"Late on Friday afternoon (after 2:30 p.m.), I was working in my room struggling with a tricky optimization problem, namely how to position my chair and an open drawer of my desk to achieve the most comfortable position in which to rest while scanning The Annals of Statistics. A knock on my door disturbed my intense concentration (this is not a euphemism for it woke me up!), and into the room entered a person whom I had never met before, but who proceeded to inform me that he had a "small problem" with which he needed some statistical help. Ever eager to please I asked him to explain his problem, and was almost at once bombarded with words and phrases such as "visual cortex", "arrays of implanted electrodes," "phosphenes," "ghosts," etc., etc. I eventually managed to stop the flow and reminded my client that I was a statistician not a physiologist and that he would need to explain his problems using language I could understand. Clearly this came as a surprise, as if the fact that a phosphene is a spot of light seen by a blind person with an array of electrodes implanted onto the visual cortex, when these electrodes are stimulated by radio waves across the skull, was self evident! Nevertheless, I persevered and eventually began to understand what appeared to be a very interesting problem, and we began to discuss possible approaches to its solution. All of these involved a considerable amount of work on my part and I estimated that it would take me at least 3 months before I would have any answers. "Oh that's no use" was the reply, "I need the results in the next 2 weeks so that I can finish writing up my M.D. thesis!" (I later discovered that the data had taken 5 years to collect and yes, the visit to me was the first to a statistician!)"
The quoted story was published by E.S. Everitt in Statistical Science, 1987, vol. 2, No. 2, p 107-134. It shows that successful consulting generally requires both statistical skills and interpersonal skills. The statistician's contribution goes beyond just cranking out numbers. Statistical consulting is defined as the collaboration of a statistician with another professional for the purpose of devising solutions to research problems. Statisticians in a university often have a heavy consulting load and a strenuous teaching schedule. Clients come to a first consultation with varied expectations about what statistical consultants do. The most common roles that consultants are expected to assume are those of helper, leader, data-blesser, collaborator, and teacher. These roles and others present several challenges.
This anecdote emphasizes the significant service I provide to helping each project through to its completion. Each of my clients (or collaborators) and I pool our talents and expertise to produce research better than that which would have occurred in the absence of statistical consultation. My involvement with several projects in cooperation with UNLV's biology department, geology department, and UN School of Medicine characterizes my consultant role at UNLV as unique in scope, important, and valuable.