Research Article

The Relevance of Statistical Knowledge for Teaching to Health Care Professionals: Reflections on a COVID-19 Press Briefing

Published: 2021-1-2

Journal: Journal of Statistics and Data Science Education

DOI: 10.1080/10691898.2020.1851160

Abstract

## Abstract Statistics is an important component of the knowledge base for health care professionals. In this essay, it is argued that statistical knowledge for teaching (SKT) should be considered an important component of their preparation as well. Health care professionals often must help others understand the statistical basis for recommendations they make. A COVID-19 press briefing is used to illustrate the need for SKT when making high-stakes recommendations related to public health. It is conjectured that efforts to educate the public during the press briefing would have been enhanced if the presenters had deeper knowledge of the general public’s common statistical thinking patterns, the typical statistics curriculum experienced by members of their audience, and contemporary tools for teaching statistics. The importance of such knowledge to support smaller-scale individual interactions is also discussed. A call for SKT-centered partnerships between educational researchers and medical researchers is made; such partnerships could be mutually beneficial to the development of both fields and to society at large. **Keywords:** Dynamic statistics software, Medical research, Pedagogical content knowledge, Statistical knowledge for teaching, Statistical thinking ## 1 Introduction It is widely accepted that statistics is an important part of the knowledge base for those in the health professions (e.g., World Federation for Medical Education Citation2015; Oster and Enders Citation2018). Statistical knowledge allows health care professionals to understand research literature and conduct their own studies. Although the specific statistical concepts to be addressed and appropriate pedagogical approaches are at times under-specified in standards for medical education (Hayat Citation2014), few would argue against requiring those entering the health professions to study statistics. This essay focuses on a different type of knowledge that health care professionals should have: statistical knowledge for teaching (SKT). SKT is not equivalent to statistical knowledge. Concisely stated, SKT is knowledge that allows one to make statistics comprehensible to others (Groth Citation2013). Health care professionals need to help people make evidence-based decisions about courses of action for preserving and improving their well-being (Nelson et al. Citation2002), though their ability and confidence to do so are sometimes questionable (Rao Citation2008). Doctors and nurses frequently make data-based recommendations during private interactions with individuals; government officials charged with preserving public health are sometimes called upon to do so on a large scale in public. This educational responsibility was especially visible during the COVID-19 pandemic. This essay argues that having statistical knowledge alone is not sufficient to teach the public about the evidential basis for recommended health practices. It is also important to have knowledge of effective teaching practices that account for how people generally interpret statistics. The essay begins with a brief description of the nature of SKT and how it overlaps, yet differs from, statistical and mathematical knowledge. Then, the notion of SKT is used to reflect upon a widely viewed press briefing in which two prominent health care professionals in the United States tried to explain the statistical basis for COVID-19 mitigation measures and treatments at a critical point in the pandemic timeline. The major statistical concepts they dealt with during the press briefing are discussed, and conjectures about how their attempt to educate the public may have been enhanced by deeper SKT are made. Broader implications about the need for SKT among health care professionals around the world and strategies for meeting the need are then considered. ## 2 The Nature of SKT SKT encompasses more than just statistics subject matter knowledge; it also includes pedagogical content knowledge (Groth Citation2013). Shulman (Citation1987) characterized pedagogical content knowledge as the “special amalgam of content and pedagogy that is uniquely the province of teachers, their own special form of professional understanding” (p. 8). Having subject matter expertise alone does not make one an effective teacher; good statisticians may or may not have pedagogical skill. Pedagogical content knowledge has been studied extensively within the context of teachers in conventional classroom settings (Depaepe, Verschaffel, and Kelchtermans Citation2013); the present article extends the consideration of pedagogical content knowledge to health care professionals charged with making evidence-based recommendations to the public. Hill, Ball, and Schilling (Citation2008) discussed three essential components of pedagogical content knowledge: knowledge of content and students, knowledge of content and teaching, and curriculum knowledge. These three components have been used widely in research on SKT (Langrall et al. Citation2017). Knowledge of content and students involves knowing how individuals commonly understand, and misunderstand, statistics content. Such knowledge supports knowledge of content and teaching, which allows for the selection of pedagogical strategies that leverage existing student understandings and address their common misunderstandings. A third component of pedagogical content knowledge, curriculum knowledge, provides understanding of the broad frameworks and learning progressions that generally undergird the teaching of statistics. Along with being more than just statistical subject matter knowledge, SKT is also more than just mathematical knowledge. Undoubtedly, a great deal of mathematical knowledge is needed to teach statistics. However, considerable amounts of nonmathematical knowledge are needed as well. For example, those who teach statistics must know how to distinguish between deterministic and stochastic questions, construct surveys, design studies, and interpret results in context (Groth Citation2007). Mathematical knowledge is necessary, but not sufficient, for such tasks. Many statistical tasks are highly context-dependent and not reducible to deterministic algorithms and conclusions. At its core, statistics is the study of variation and uncertainty. This stands in contrast to mathematics, in which deductive reasoning can lead to definitive proofs. Mathematical tools are helpful for statistical tasks, yet we cannot expect them to produce conclusions as certain as those of mathematical proofs in most statistical contexts. Reconciling certainty in mathematics with uncertainty in statistics is a nontrivial task. Knowing how to help the public distinguish between statistical and mathematical conclusions during a time of anxiety such as the COVID-19 pandemic is especially challenging, as people naturally seek unequivocal answers to questions that cannot be answered with complete certainty. ## 3 SKT in the Context of a COVID-19 Press Briefing In the United States, at the outset of the pandemic, Dr. Deborah Birx and Dr. Anthony Fauci were charged with explaining the basis for government-led COVID-19 mitigation and efforts to develop treatments and vaccines. On March 31, 2020, the two doctors employed statistical tools and arguments during a press briefing to do so. The press briefing came at a critical time, as the United States had recorded just a small percentage of its eventual deaths from the virus (approximately 4000 of what would grow to be over 250,000), and audiences for the daily briefings from the White House were averaging approximately 8 million viewers (Grynbaum Citation2020). A primary goal of the briefing was to argue in favor of extending existing mitigation measures, which included stay-at-home orders, business closures, and maintaining social distance in public. Video of the press briefing was posted online afterward (e.g., https://www.c-span.org/video/?470841-1/white-house-warns-upcoming-painful-weeks&live=). Quotes in this essay were drawn from the government video transcript (available from https://www.whitehouse.gov/briefings-statements/remarks-president-trump-vice-president-pence-members-coronavirus-task-force-press-briefing-15/). The statistical subject matter of the press briefing included models, disaggregation of data, and experimental study design. Next, a summary of how Drs. Birx and Fauci dealt with each of these statistical ideas during the press briefing is given, followed by conjectures about how deeper SKT related to the three main statistical ideas may have enhanced their efforts to teach the public. ### 3.1 Press Briefing Introductory Presentation Summary Dr. Birx began her portion of the press briefing by projecting the slide shown in Figure 1. Two curves appeared on the slide: one to illustrate the spread of COVID-19 without mitigation measures, and another to illustrate its spread with mitigation measures fully in place. She explained that the curves were based on modeling done by “five or six international and domestic modelers from Harvard, from Columbia, from Northeastern, from Imperial who helped us tremendously.” In reference to the curve on the far left in Figure 1, Dr. Birx stated, “In their estimates, they had between 1.5 million and 2.2 million people in the United States succumbing to this virus without mitigation.” She explained that the curve on the far right showed projected deaths if mitigation measures such as social distancing and hand-washing were fully implemented, saying, “what an extraordinary thing this could be if every American followed these, and it takes us to that stippled mountain that’s much lower—a hill, actually—down to 100,000 to 200,000 deaths, which is still way too much.” ... [the content continues in a similar detailed manner] ...

Faculty Members

  • Randall E. Groth - Seidel School of Education, Salisbury University, Salisbury, MD

Themes

  • Education and Training of Health Care Professionals
  • Health Literacy
  • Statistical Knowledge for Teaching (SKT)
  • Statistical Thinking
  • Pedagogical Content Knowledge
  • Public Health Communication

Categories

Download Article