For the TLDR group, just read the last paragraph.
What I have always loved about statistics is the breadth of topics I get to learn about. It's what drew me into statistics in the first place - when I had my first undergraduate statistics course and heard my professor talking about the varied projects he worked on. What I have always disliked about statistics are the consequences that often accompany the lack of ownership in those same types of consulting projects. With that lack of ownership comes someone else dictating a budget for my time and making decisions about whether they will take my advice or not. It has also meant that when financial times get tough, allocation for statistical consulting is one of the first things that has been cut from project budgets, which ends up being exchanged with the client muddling through the statistics on their own the best they can, which often results in statistical errors that they never even know about (I know this actually happens a lot because I have had numerous occasions to get added to a graduate student's Masters or PhD committee after such work has been done, or added to a private industry project in its late stages).
I believe the lack of ownership and the budget constraint issues highlight one of the major underlying factors that dictates when my involvement is more of a consulting role or a collaborator role. As I previously mentioned, I have worked in industry, currently work as both a statistical consultant and collaborator in academics, and have a moderately active private consulting business. As many people have pointed out, consulting involves collaboration. However, in some of my projects I distinguish myself as a consultant and on some I distinguish myself as a collaborator, though there is not a precisely defined line distinguishing the two. Rather, it is a gradient that is composed of several scales, including level of ownership, time investment, time of inclusion, and tasks I participate in, and very possibly other scales.
I think it is interesting to consider what my clients/collaborators would say my role has been. There is no doubt that on some projects I would be called a collaborator, while on others I would be called a consultant, but I'm pretty sure there would be high consistency among members of a project as to how they viewed me, which indicates to me that somehow we have had some commonality in our perspective of what constituted collaboration and what constituted consulting. In more recent years at the university, I have been very upfront about what role I'm willing to take when asked if I would work on a project. Virtually never does anyone ask if I will be one or the other, but I am almost always explicit about what role I'm willing to take on, and then they can choose if that is agreeable to them. It wasn't always like this in my university work environment, but our role as consultants was not valued much by our peers or our evaluators, so we have moved to a different level of participation for many of our projects. I still take on the role of consultant occasionally, but it is usually because a project doesn't interest me enough to devote more time to it or because I just don't have the time, but I have to balance this choice with the knowledge that it will not be valued much in my evaluation.
The move to collaborator has been motivated by the values of my evaluators and my non-statistician peers. It is extremely important to note that this does not reflect my personal values on consulting and collaboration. From a personal satisfaction perspective, I generally value consulting more because, for a fixed amount of time, I tend to learn more about statistics, I get to meet more people, and I get to see a greater variety of projects. And monetarily, I have tended to earn more money in consulting roles. But my current employer doesn't value consulting very much, despite exceptional efforts to educate our administration and our peers of the value of consulting, which has included letter-writing on our behalf by our clients. I understand if this offends any of my fellow statisticians reading this because it has been a struggle for the members of our statistics group as well, yet it is the way things are. But I can also understand that it is the broader culture of our work environment that has led to this, and occasional it does in industry as well.
Budgeting and the way time allocations are governed has a lot to do with this. Commonly at the university, occasionally in industry, and extremely rarely in private consulting have I seen almost no budget limitations on the statistical work done. Minimal budget limitations have allowed me to use my judgment on how much I invest in a project. More commonly in industry and virtually always in private consulting, there have been budget constraints and, consequently, the project owner (or funder) rarely anticipates that the additional contributions I would make as a collaborator, beyond what I will make as a consultant, is worth the additional money. So the values that guide decision making in industry are often very different from what they are in academics.
While I will always consider myself a consultant at heart, my work environment often dictates which role I play, and it is different. Neither is of greater or lesser value to me, I am simply addressing the values of those who control my work environment. The point here is that we all have different work environments, and I hope that we can understand that *IN SOME WORK ENVIRONMENTS* the role of collaborator is genuinely different from the role of consultant, though mostly in where our role lands us on a variety of component scales. Though some statisticians may prefer one role over the other, there is no generally implied value judgment on the part of the statistician, it's just different. The people who control our work environments, on the other hand, often do place different values on the roles, and we will tend to cater to what those values are.
----------------------------------
David Daniel
Professor
Applied Statistics
New Mexico State University
ddaniel@nmsu.edu------Original Message------
I am out on vacation right now but saw this and wanted to reply--my point, and the point of the five WJ Dixon award winners who wrote this, was NOT to say that those who view themselves as consultants do not collaborate. The ongoing goal, which in my opinion requires much more conversation within the community, is to find terms to discuss the difference between what is described in the article as "consulting" and what is described as "collaboration". We cannot just throw out these terms and tell everybody what they now mean--and that was not the point of the article. This is a starting point, not an ending point.
I do feel this is a different issue in academia than in the private world and this requires a much longer email than I have time to write at the moment, but I hope to do so upon my return next week. I would like to explain why finding ways to discuss this is important, but it must be a discussion, not a dictation.
As someone with a foot in both the academic world and the private world, it's extremely important to me that this is discussed in a way that does not isolate anyone.
Kim
------------------------------
Kim Love
Owner and Lead Consultant
K. R. Love Quantitative Consulting & Collaboration
------------------------------