11 Writing About and With Numbers

As readers and writers engaged in professional and public problem-solving, we need numbers.  But it seems that few people who are comfortable with words are comfortable with numbers, and few people who are at home with numbers feel the same way about words.  The purpose of this chapter is to point the way to productively integrating words and numbers in the service of engaging problems that matter.

In Writing Studies, we all use words to articulate our “wicked problems” and to give voice to perspectives that might be silenced or underrepresented.  Writers use personal anecdotes and historical examples and expert testimony to express the hardships and histories of gun violence, racial profiling, health-care and educational inequities, and environmental crises in order to expand their discourse communities and to reach more stakeholders.  But establishing full credibility—ethos– often requires clear expressions of numerical reality.  Whether the numbers express simple facts, comparative statistics, or variable rates of change, professional and public writers need to become comfortable evaluating, gathering and expressing numerical data effectively toward their own rhetorical ends. And importantly, the less comfortable our reading audiences are with numbers, the more thoroughly and explicitly we as writers need to interpret and explain them.  Despite the age-old saying that “the numbers speak for themselves,” rhetorically-attuned writers know they certainly do not.  Before they can be accurately evaluated by active readers, numbers must be productively interpreted and contextually explained by focused, purpose-driven writers.

As readers, it is important to remember that while numbers and data can help us understand the reality of a situation, they can also obscure and distort problems, too:  numbers are only as honest as the human being expressing and explaining them.  We know from expressions like “There are lies, damned lies, and statistics,” that there is a history of numbers being mis-used in order to direct public attention toward or away from a problem.  Especially in this age of misinformation and social media, the ability to be numerate (that is, to be someone who can understand and work with numbers, the same way “literate” means someone who can understand and work with written words) matters more than ever. We should become habituated to evaluating statistics and numbers the way we evaluate any other source:  Is the data accurate?  Credible?  Timely? Who made the statistic?  Why? For whom? How are the terms of the data being defined?

As Professor of Sociology and Criminal Justice Joel Best warns us:

Reality is complicated, and every statistic is someone’s summary, a simplification of that complexity.  Every statistic must be created, and the process of creation always involves choices that affect the resulting number and therefore affect what we understand after the figures summarize and simplify the problem.  People who create statistics must choose definitions—they must define what it is they want to count—and they must choose their methods—the ways they will go about their counting. Those choices shape every good statistic, and every bad one. . .. Good or bad, every statistic reflects its creator’s choice (Damned Lies and Statistics, 160-161).

So, in the same way we evaluate written summaries to be sure they are accurate, inclusive, and not hiding authorial bias, we also need to evaluate statistics and other forms of quantitative data we come across, understanding that numbers are as much the product of human choices as they are descriptions of objective realities.

As writers, we need to be alert to how our own use of statistics and other numbers can ethically and effectively direct attention to our problem.  Joanna Wolfe, Professor of English, focuses on three areas where writers have the opportunity and obligation to make robust meaning when presenting quantitative data in a rhetorical situation.  She tells us to be mindful of three areas: “Pathos, ethos, and statistical argument” (Wolfe, “Rhetorical Numbers,” 458).

First, when writers translate facts into statistics, pathos comes into play.  Wolfe gives an example by comparing the emotional effect of being told that “one in fifty pregnancies by women over thirty-five will result in an abnormal fetus” to being told “there was a 97 percent likelihood that her unborn child would have no problems” (“Rhetorical Numbers” 459).  Imagine all the different ways this statistical reality could be presented, and it becomes clear that rhetorically effective writers have many choices.  The material fact is the same but the rhetorical expression of that same fact can evoke a very different appeal to pathos. So as writers, we must be intentional about the way we choose to express numerical facts to our audience, given the purpose and effect we hope to achieve.

Second, to relate data to the appeal to “ethos,” Wolfe notes the way the expression of quantitative data follows from the very human act of defining terms, of using, as she says, “contestable definitions.”  To illustrate this point, Wolfe analyzes the following statement that claims to be an objective “fact”: “‘Korea- a nation that only recently rose above third-world status- spends half what the U.S. does per student, yet comes in 13 places ahead of the U.S. on an international math assessment’” (Wolfe, “Rhetorical Numbers” 461). Here is her analysis of that statement, in which she asks several critical questions about the statement:

Acceptance of this claim requires that we agree upon several definitions, the first of which is “spending.” Is spending here being defined as raw dollars, proportion of GNP, or some other measure? If raw dollars, then it is unlikely that $1,000 spent in Korea is the same as $1,000 spent in the United States. And does spending include money spent on public education only or public and private together? Does it include postsecondary education? Similarly, the definition of “student” in this statistic is also up for debate: does this category mean the same thing in both nations? Does Korea have the same mission of universal education as the United States?  if one country has a higher drop-out (or weed-out) rate the factor will influence overall score performance. (Wolfe, “Rhetorical Numbers,” 461-2)

Wolfe’s critical analysis of this supposedly “factual” claim helps us see that just as every statistic is a summary that we need to evaluate for accuracy and inclusion, every policy statement that uses quantitative data is also a human product that needs to be evaluated with an eye toward the credibility of its maker.  Are we all in agreement about the definitions used by the author?  Is the author counting what we assume they’re counting?  Is the author using numbers to make fair comparisons?

Wolfe’s third point, following her attention to pathos and ethos, is about the role of “rhetorical invention and arrangement in quantitative arguments” (Wolfe, “Rhetorical Numbers,” 459). In other words, raw data expresses socially meaningful realities only when rearranged and grouped according to selected metrics like gender or ethnicity. To illustrate her point about the importance of rhetorical “arrangement,” Wolfe includes a long, randomly arranged list of fictitious test scores that seems to express no meaning or significance.  But when the raw data is re-arranged using metrics of gender, ethnicity, and school, patterns of achievement—and non-achieivement– emerge.  Only once readers have been directed to note these patterns of achievement and struggle by a deliberate rearranging of the quantitative data, can appropriate interventions begin. The ability to make this kind of arrangement is how writers of quantitative arguments foreground the stories they want to tell with their data.

Using numbers well allows upper-level writers to experience their power as researchers within and across discourse communities.  To reemphasize Best’s and Wolfe’s points with some current examples related to recent student topics in WTNG 225, consider the following:

• We can use words that appeal to audience’s pathos to persuade them to feel a certain way about the numbers we present. In this example that might interest students studying Construction Management or Sustainability, the first sentence simply states a numerical fact about cement, but in the second sentence, the writer from Statista has included words like “massively” and “more than doubled” to create troubling images about the impact of the fact in the readers’ mind:  “Global emissions from the manufacture of cement stood at 1.6 billion metric tons of carbon dioxide (MtCO₂) in 2022. Emissions from cement production have increased massively since the 1960s, and have more than doubled since the turn of the century” (Statista, Carbon Dioxide Emissions).  Without the second sentence, can the first sentence alert you to a problem?
• We can be careful about clearly understanding (as readers) and defining (as writers) the concepts that provide context for the numbers we use. The following example about college students and mental health provides a good opportunity to do some careful math and to consider the importance of defining what is meant by a “mental health problem”: “The majority of college students (more than 60 percent) meet the criteria for at least one mental health problem—a nearly 50 percent increase since 2013, according to the Healthy Minds survey, published earlier this month” (NEA Today, March 29, 2023). How would your understanding of the problem be shaped by the writer’s choice about whether or not to specify what a “mental health problem” is?
• And eventually, writers engaged in original research may be empowered to choose the metrics they use to organize and arrange the raw data they gather or generate. For example, the researchers at the data-rich website Prison Policy Initiative , as we might expect, show charts on demographics of incarcerated people.  But they also have extensive reports on incarcerated people’s health conditions and their access to health care. (Prison Policy Initiative, “Chronic Punishment). It’s worth thinking about why, when, and to what affect this organizing metric was included by researchers on mass incarceration.  How might this new metric change the stakeholders in the discourse community comprising Criminal Justice?

In conclusion, writing about problems that matter will almost certainly require writing about and with numbers. Numbers—quantitative data– allow writers to inform readers about the extent, scope and scale of any problem’s parameters as well as the rate of the harm being done. In fact, it is ONLY through our sustained attention to the quantitative data that we understand how big any problem is, how fast it is growing, and, ultimately, whom it affects.  As problems become increasingly complex, it is more important than ever that as readers and writers, we become “numerate” —literate about numbers and their contexts. What wicked problem does not require nuanced engagement with quantitative data? As our engagement with a problem deepens, we need to drill down to increasingly complex measurements (of something, or many things) taken over time.  So, at the most basic level, all citizens need to become accustomed to reading quantitative data in critically effective ways, and all writers need to become practiced at presenting quantitative data in rhetorically effective ways. Whether writing a proposal, report, or white paper for a workplace environment; a long-form essay for a public audience; or a research paper in an academic context, effective writers need to evaluate and explain the numbers they’re using to make the strongest possible connection to the audience of readers.

Works Cited

Best, Joel. Damned Lies and Statistics:  Untangling Numbers from the Media, Politicians, and Activists.   Berkeley and Los Angeles, CA:  University of California Press, 2001.

NEA Today, “The Mental Health Crisis on College Campuses,” March 20, 2023.  Retrieved January 9, 2024.Prison Policy Initiative, “Chronic Punishment: The unmet health needs of people in state prisons”, June 2022.  Retrieved January 9, 2024.

Statista, “Carbon Dioxide Emissions from the Manufacture of Cement Worldwide from 1960-2022.” Retrieved January 9, 2024.

Wolfe, Joanne “Rhetorical Numbers: A Case for Quantitative Writing in the Composition Classroom.” College Composition and Communication, February 2010, Vol. 61, No. 3. pp. 452-475.

For further reading:

Edward Tufte is Professor Emeritus of Political Science, Statistics, and Computer Science at Yale.  He is a pioneer in the field of data visualization.  Any student interested in the intersections of quantitative data, fine arts, social justice, and technologies will enjoy his website, https://www.edwardtufte.com/tufte/

Joanne Wolfe, whose work is vital to this chapter, is the author of several books and articles about the importance of including quantitative rhetoric in writing courses, especially for students in Engineering and other highly technical fields.  See especially “Teaching Students to Focus on the Data in Data Visualization.” Journal of Business and Technical Communication 2015, Vol. 29(3) 344-359.