Collaboration and Communication: Challenges for Statisticians

Marcey Hoover

Data Torturing and the Misuse of Simple Statistics

Abstract: Statistics are often misused by not applying them in the context of statistical thinking that focuses on processes, variation, and data. The result of this misuse may be “data torturing” or going beyond reasonable interpretation of the facts due to a misunderstanding of the processes creating the data or the misinterpretation of variability in the data. The consequences of data torturing are often amplified when well-intentioned technical experts use simple statistics to communicate results to decision makers. This talk will provide examples where the communication of simple statistics, in the absence of statistical thinking, provides deceptive results by not adequately representing the underlying process and variability. The examples are drawn from a career in national security and demonstrate how even in very complex and high consequence situations, statistics can be misleading if not communicated with proper context. Statisticians can play a key role in mitigating the risks associated with data torturing by assuring that the concepts of statistical thinking are applied and properly communicated.

Domenic Reda

The Statistician as a Clinical Research Team Facilitator

Abstract: Clinical researchers typically develop an idea for a clinical trial on their own or in a working group of clinical researchers.  After the study design is finalized, the team may approach a statistician for help with the sample size calculation and to develop a statistical analysis plan.  A more desirable arrangement is for the statistician to be involved at earlier stages in the development of the research plan, where they can contribute to discussions on trial feasibility, potential biases in the design, and key components of the design such as trial eligibility criteria, choice of outcome measures and treatment groups. This talk will discuss an even more fundamental role for the statistician, where the team of clinical investigators have disparate ideas on what questions the study should ask and the team is unable to arrive at a consensus.  One such example is the design of a phase II trial to evaluate the potential effectiveness of cannabis as a treatment for neuropathic pain for people with diabetic neuropathy. A group of neurologists, psychiatrists and clinical psychologists formed a working group to propose a study in this area.  At least four different studies were proposed and the group was unable to arrive at a consensus.  I was approached to help facilitate the group, evaluate the various designs and propose a design that all could agree on.  The group was able to achieve consensus, complete the study proposal and received a fundable score from peer review.

Advancing Global Development:  A Role for Statisticians

Abstract: The international community has identified important goals for global development, including eradicating poverty and hunger, mitigating and adapting to climate change, and otherwise improving the lives and well-being of populations.  Data and statistics are essential to guide policies and programs to achieve these goals.  However, the international bodies, governments, non-governmental organizations, and communities working towards development goals often lack the statistical capacity to use data and evidence for decision-making, creating an important role for those engaged in translational statistics.  Collaboration is key with these partners, and the limited statistical capacity increases the burden on statisticians to support communication of findings to diverse audiences and translation into practice.  Additional efforts are needed to work across cultures and languages and to navigate and reduce inequities in training, opportunity, and resources.  Statisticians’ ethical responsibilities also increase when a partner lacks background to provide oversight, and there can be tensions between building statistical capacity and providing statistical consulting or collaboration.  This talk will provide examples of statistical applications in global development and the challenges described above as well as suggestions for guidance and training of statisticians interested in global development work.

John Newell

Translational Statistics: Surviving the Test of Time

Abstract: Translational Medicine promotes the convergence of basic and clinical research disciplines and the transfer of knowledge on the benefits and risks of therapies. In an analogous fashion we proposed the concept of Translational Statistics both as a statistic used to report results of an analysis of data and as an approach to communicate the results of an analysis in an accurate and accessible manner to diverse audiences^. The use of appropriate graphics is central to all aspects of translation. For example, the graphical representation of data is necessary to provide a way of assessing at least parts of any assumed statistical model before engaging in formal analysis and to aid in the presentation and communication of results and conclusions.