Statistical sciences sometimes have difficulty getting their voice heard in the broad spectrum of physical, medical and social science, and very likely we have all experienced the necessity to explain the role of statistical thinking to administrators, supervisors, granting agencies, and colleagues from other areas of specialization.In Rodney Dangerfield terms: “we don’t get no respect”.
We believe that the situation has changed greatly for the better over the past ten or fifteen years, and that the respect we are looking for is no longer elusive. One of us (Reid) has personally experienced a very pleasant and welcome respect for the importance of statistics in several activities related to interdisciplinary work.Serving on the review committee of the Health Effects Institute, she was consistently impressed with the statistical level of the discussions around the table, and the importance attached to statistical thinking. In a forestry project under the leadership of CD and two of her colleagues, because of challenges faced with pest and fire concerns, scientists have been urgently vocal on the need for the development of novel statistical methods for forest ecology and fire management, and have been tremendously supportive that national funding be directed toward such methodological statistical developments. Colleagues across both our campuses have made it clear that they need and value statistical thinking, and are very willing to engage in and fund collaborations with statisticians. In 2003 in Canada a group of statisticians established the National Program on Complex Data Structures, with the explicit goal of establishing interdisciplinary collaborations in which statisticians played a leadership role, and this program was very successful.
Our concern now is that as a community we are, ourselves, creating roadblocks for further advances of our discipline. We have a strong tradition of skepticism, and arguably a healthy skepticism, that generally ensures we are very cautious in embracing the ‘next big thing’. We also have a strong tradition of argumentation, and some important statistical advances have come out of these arguments.
Quite recently two dramatic examples of this surfaced in Canada, in grant reviews at one of our major granting councils, NSERC. In one case, an interdisciplinary proposal was favorably reviewed by international and interdisciplinary committees, and heavily criticized by a panel of domestic statistical reviewers. In the other, individual grantees were compared across disciplines for special funding, and statistical reviews of leading applicants were simply much more reserved than those in other disciplines. With the rising trend to engage in collaborative work and seek funding from interdisciplinary panels, statistical sciences lose out to other disciplines because their grant proposals are not well supported by statistical reviewers.This is an embarrassment for our community in an interdisciplinary world, but more importantly it means we don’t get the money. In a time of tight budgets and ongoing pressure in every field of science our scientific skepticism and willingness to criticize is hampering our progress.
We all value statistical science and want to see it valued as an important part of the scientific enterprise.However, our scientific skepticism has spilled over into an ineffectual skepticism of statistical scientists, and it is time to have a discussion throughout our community on these issues. What can we do to change our culture to one where we are pleased when our colleagues are successful? To one where we recognize that a broad spectrum of statistical activity is valuable and important? To one where we are open to and indeed, get excited about, new ideas diametrically opposed to our own, even if we (secretly) feel that our own ideas are better? To one where we finally realize that by not supporting each other with enthusiasm, we are indeed shooting ourselves in the foot.
-----