Turning on the Historian's Macroscope: A Call to Foreground the Teaching and Learning of Data Visualizations in World History Education

Tamara L. Shreiner

     In their recent book about exploring "big data" in history, Shawn Graham, Ian Milligan, and Scott Weingart describe a tool for historical analysis that they call a "historian's macroscope."¹ "A macroscope," the authors explain, "is a bit like a microscope or a telescope, but instead of allowing you to see things that are small or far away, the macroscope makes it easier to grasp the incredibly large."² According to Graham, Milligan, and Weingart's description, such a tool is essential in world history. While all historians are experts in looking at information from the past through a microscopic lens, world historians must regularly transcend traditional boundaries of time and space, connecting that which they see through their microscopes to larger patterns and trends at the global level.³ As Graham, Milligan, and Weingart explain, the historian's macroscope compresses large amounts of data to make "once-obscure patterns and relationships become clear," often revealing such patterns through "data visualizations in lieu of direct images."⁴

     It is no surprise, then, that world history texts are replete with data visualizations—maps, graphs, and charts that are products of the historian's macroscope and depict trends and relationships impossible to see with the naked eye. Data visualizations are everywhere in world historical trade books, journals, and digital history projects, and they play a prominent role in the textbooks that often dominate world history instruction.  Such data visualizations are not mere decorations intended to break up the verbal narrative; more often than not they elaborate upon or provide visual evidence for the written arguments and explanations surrounding them. ⁵ It is therefore important that students develop the skills to work with data visualizations in world history.

     This article makes a case for foregrounding data literacy teaching and learning in the field of world history. While the prevalence of data visualizations in world history texts may come as little surprise to anyone familiar with the field, the effective teaching and learning of data visualizations in service of world historical understanding deserves greater attention. As several researchers have argued, data visualizations can be complex, and students face a number of difficulties in analyzing, interpreting, and using them.⁶ Therefore, this article is an attempt to illuminate both the affordances of data visualizations in teaching and learning world history, and the challenges they pose.⁷

The Role of Data Visualizations in World History

     Though data visualizations come in seemingly endless variety, they can be categorized into six basic types: simple text, tables, charts, graphs, network graphs, and geospatial maps. Simple text visualizations display a quantity as a number or percentage, often boldly and in color. Tables are row-by-column matrices, with cells that can be color-coded or sorted and can contain graphic symbols or icons. Charts are visualizations with no inherent reference systems, such as pie charts or word clouds. Graphs, on the other hand, have a well-defined reference system, such as horizontal or vertical axes on which data are plotted. There are sequential graphs that show movement, causal relations, or organization, such as timelines and time series, as well as quantitative graphs that have conventional presentations of numerical data on x- and y-axes, such as line graphs, bar graphs, and scatter plots. Network graphs are their own unique type of data visualization and include taxonomies, social networks, or migration flows. Finally, geospatial maps, a common form of data visualization, use a latitude and longitude reference system overlaid with physical or political markers. Data visualizations can further be grouped into larger taxonomic categories based on the types of questions they address: temporal data visualizations answer "when" questions; geospatial answer "where" questions; topical answer "what" questions; and network answer "with whom" questions. There are also spatiotemporal data visualizations that combine geospatial and temporal data, answering both "where" and "when" questions.⁸

     Maps are likely the most common data visualization a student will encounter when studying world history. For example, if you were to peruse the popular world history curriculum website, World History for Us All, you would find that 50 of the site's 75 units of instruction contain at least one map, adding up to 211 maps overall.⁹ Likewise, in middle school world history textbooks, maps account for approximately 72% of all data visualizations, and in high school world history textbooks, they account for 59% of all data visualizations. Compare this with U.S. history textbooks, where maps account for 58% of all data visualizations in middle school, and 49% in high school.¹⁰

     Although using the term "maps" may imply a monolithic collection of geospatial visualizations, in fact, the variety of maps students will find in world history is striking. They include choropleth maps, connection maps, and flowmaps, and use a wide array of scales, from large-scale maps of states or regions, to small-scale maps of the entire globe. Perhaps more striking is the prevalence of spatiotemporal maps—complex, multi-layered maps with data indicating change, growth, or movement over time. In fact, spatiotemporal maps are the most frequent kind of visualizations students encounter in their world history textbooks, accounting for 42% of all visualizations.¹¹

     The prevalence of maps in world history texts is by no means surprising. To be sure, maps have played an important part in the human past and are, therefore, often depicted as part of the world historical narrative. Way-finding maps, for example, have led travelers, explorers, and conquerors across waterways and to new lands, and facilitated the spread of ideas, technologies, foods, diseases, and people. When viewed as primary source evidence, maps can reveal to historians the habits, thoughts, and perspectives of the people who created them, helping us understand how people of the past saw the world, where they had gone, and where they thought they could go.¹²

     Yet maps are also prevalent in world history texts because they are tools of inquiry and communication for world historians. They can serve as illustrations for arguments and explanations about the global past, visualizing spatial relationships often hidden in written text and graphs. Historians use them to show how people have moved over space and time, how societies have grown or dissolved, how diseases or languages spread, and how resources and people are distributed. Because maps can show both the entire world and parts of the world at the same time, they allow world historians to illustrate the intellectual shifts they are making as they move across scales and weave a narrative about changes at both local and global levels.¹³ Consider, for example, the seminal world historical work, The Human Web, by J.R. McNeill and William H. McNeill. They used over a dozen maps in an otherwise visual-free text to show historical phenomena such as how agriculture was invented in separate parts of the world, or how Muslim empires expanded from the 7^th to the 10^th century CE.¹⁴ Further, imagine geospatial-based arguments like that of Janet Abu-Lughod in Before European Hegemony: The World System A.D. 1250–1350 without the use of maps to illustrate the world system at the heart of her thesis.¹⁵

     Other data visualizations, such as tables and charts, are also important products of the world historian's macroscope. They aggregate and compress data until obscure or invisible patterns and relationships become clear, making it easier to grasp the incredibly large. World historians can use them to make connections between local and global, to make comparisons between regions or time periods, or to illuminate global patterns. They can use them to test hypotheses about large data sets or to provide evidence for historical interpretations about how and why past changes occurred.¹⁶

     Indeed, as with maps, recent research shows that temporal and topical graphs and charts are prevalent in middle and high school world history textbooks, providing information that is not in the written narrative.¹⁷ World History for Us All includes at least one data visualization like a table, chart, or graph in 36 of its 75 units, with a total of 177 across all lessons. A recent review of world history articles from The Journal of World History revealed that over ten years, 13% of the original articles in the journal contained tables, graphs, or charts to help support authors' arguments. Journal authors used such data visualizations to make original arguments about historical change, to compare changes or consequences across regions, or to critique other historians' interpretations and uses of comparative data.¹⁸

     Given the role that data visualizations play in world history, students must be equipped to make sense of maps, graphs, and charts and to construct, interpret, and critically analyze world historical arguments that use them. Recent research even suggests that working with data visualizations in history has the potential to enhance students' historical reasoning skills.¹⁹ In particular, data visualizations can supplement written text used to answer historical questions and help students better contextualize a historical situation under study, which multiple researchers have identified as a challenging aspect of historical reasoning.²⁰ Data visualizations can also help student reason about historical questions by prompting them to consider causal factors related to a historical event, or by illustrating change over time that might not be readily apparent in written narratives.²¹

     However, the ability to effectively read and analyze data visualizations does not come naturally to students. In fact, research indicates that students are likely to entirely ignore data visualizations embedded in texts.²² Even when students do attend to data visualizations, reading and drawing inferences from them can be challenging. In what follows, I summarize recent research that highlights students' challenges in making sense of data visualizations in history.

Students' Challenges with Data Visualization Comprehension

     Although it is often assumed that data visualizations are easy to understand, several researchers  have pointed out the complexity in reading them, arguing that the ability to draw correct inferences from graphs is mediated by a number of factors.²³ First, readers must encode the full array of graphical elements embedded in a data visualization and identify its important visual elements, such as the shape and directions of a line, or numbers on an axis. Readers must then relate the visual elements to the conceptual relations that they represent, mapping between the elements and their meaning. For example, a student would need to recognize that a sloped line indicates an accelerating relationship. Finally, a reader must be able to make associations between the visual elements and their referents, such as recognizing that a graph represents population changes over time because those elements are shown, respectively, on y- and x-axes.²⁴ Similarly, reasoning with maps requires that people recognize the symbol system used to communicate data on the map, the relationships among the symbols, and the way that the map is layered with physical, political, or historical data. Understanding the system used to communicate information on a map allows people to make meaningful connections among its different layers, draw inferences, and reason about a question or problem.²⁵ Any gaps in a student's understanding of common graphical conventions, conventions of a specific type of data visualization, or content related to the data can lead to problems in comprehension of a data visualization. Such difficulty makes it unlikely that students will be able to draw inferences and integrate information across verbal and visual modes, which is what they need to do in order to fully comprehend a text and use it for a particular task, such as reasoning about a historical question.

     To be sure, a recent study of middle and high school students' uses of data visualizations in history textbooks showed that a student's inability to break down a data visualization into all of its component parts can hinder their ability to draw inferences from the data visualization. To begin, students in the study missed many of the visual features representing information in the data visualizations that they were reading.  On average, middle school students missed more than half of the visual features, while high school student missed just less than half. Students' lack of attention to all the visual features in a data visualization proved detrimental to their ability to grasp its full meaning. When attempting to use data visualizations to reason about a historical question, the number of visual features students identified was positively and significantly correlated to the number of facts that they stated in relation to the data visualization. In turn, the number of facts that students stated was positively and significantly correlated to the number of historical reasoning factors (e.g., contextualization, causation) that emerged in their responses to the question.  Furthermore, students' abilities to draw inferences from data visualizations were stymied when they lacked knowledge of the graphical conventions in the data visualizations they were attempting to read. For example, when faced with a complex data visualization that challenged their skill level—a spatiotemporal map typical of those in world history textbooks—about half of high school students in the study struggled to make sense of it, either "giving up" or talking about how confusing it was.²⁶

     All of this research points to a need to explicitly teach data visualizations in world history education. Though we currently lack research specific to history teachers' work with data visualizations in classrooms, extant research in other fields suggests that teachers commonly do little more than point to data visualizations during instruction.²⁷ This lack of explicit instruction could have unfortunate consequences for world history students. They will likely face data visualizations embedded in reading materials with increasing frequency as they move through school, such as from middle school to high school, and the data visualizations they are expected to read will become increasingly complex. Indeed, by the time they reach high school world history students will most likely encounter an abundance of complex data visualizations because of the prevalence of multi-layered spatiotemporal maps in world history textbooks alone—not to mention those they might find in additional online and print curriculum materials.²⁸ The next section begins to paint a picture of what we might do in world history classrooms in order to help students acquire the skills that they need.

Teaching Data Visualizations in World History

     At the most basic level, students must develop what scholars have referred to as "concepts of graphics." ²⁹  They must begin to pay closer attention to the data visualizations that are everywhere in the print and online materials they use in world history, and recognize that data visualizations are important representations of information in the discipline of history. Students must further understand that data visualizations have been created with intentionality by illustrators and authors, who are trying to convey a particular argument or explanation, and who may have biases that will influence the variables they use and the story they try to tell. Finally, students must recognize that data visualizations have relevance to surrounding information, often extending that which is found in surrounding verbal text.³⁰ The simplest way to help students develop concepts of graphics like data visualizations is for teachers to continuously point out both verbal and visual elements in informational texts. More involved, student-centered exercises are also helpful.  One exercise I have often used with students involves asking them to randomly choose a chapter from a school textbook and analyze the data visualizations contained therein. I ask them to write about what information each data visualization is communicating and how it is related to the main body of verbal text. This allows students to discover the role of data visualizations in texts for themselves, and to recognize that they typically provide extensional information. Another activity I have used to help both high school and college students recognize the communicative power of data visualizations is to have them create narratives or arguments that incorporate multiple data visualizations. They can do this with Word documents, or by using online discussion or blogging platforms.

     It is also essential that teachers are deliberate in how they introduce students to data visualizations for analysis. For example, some researchers have suggested that in order to avoid overwhelming students, teachers should teach about bar graphs before pie charts, and pie charts before line graphs, and teach about maps with only one or two layers before they teach about maps with multiple layers.³¹ Regardless of the specific order of instruction, it is important to be cognizant of the complexity of data visualizations we are asking students to interpret, whether they are those in the readings we assign or in the presentations we give in class. In addition, no matter the kind of data visualization, students require instruction that helps them develop skills in deconstructing maps, graphs, and charts so that they can identify all of its visual elements. One way to do this is by asking students to construct their own data visualizations so that they become aware of all the elements required to make a data visualization complete and informative. However, it is also important to help them parse data visualizations that have been produced by others. Missing visual elements of a data visualization is similar to missing phrases or sentences in paragraph—ignoring the parts can change the meaning of the whole. Yet, it is not enough to just see the visual elements. Students must also be able to make meaning from connecting the various elements, learning how to identify the facts conveyed by the data visualization. Then, they must use their prior knowledge, as well as relevant surrounding information, to draw reasonable inferences from the data visualization and interpret how the author is using the data to support or illustrate an argument or explanation.³² At the same time, students need to question the premises and choices that underlie data visualizations, evaluating the strength of the explanations or arguments that are embedded within them, and identifying features or information that were left out.³³ Below is a framework for analysis of data visualizations in history that I have provided as a tool for students in my classes. It indicates what students should be able to do and the kinds of questions they should be prepared to ask of data visualizations.  Many of these questions are similar to those we typically associate with historical thinking, so incorporating them into classroom instruction only further enforces habits of mind we want our students to develop in world history.

     One danger in the teaching and learning of data visualizations, particularly as part of the discipline of history, is losing sight of the human elements underlying the data. As Graham, Milligan, and Weingart indicate, macroscopes and the data visualizations they produce compress data in order to help us see patterns otherwise invisible to human perception. Yet, history is a human enterprise, so we must always maintain a connection to the human elements at the heart of the discipline. Teachers might therefore keep in mind two considerations for humanizing data for students. The first is to humanize the creation and presentation of data. Simply stated, data can lie. A truly data literate person understands the ways that data can be manipulated and therefore takes account of the fact that a person from within a particular context and with a particular purpose created any given data visualization. Sourcing a data visualization is a good place for students to start, paying attention to who created it, when they created it, and for what audience and purpose. Students can then consider the choices that the creator of data visualization has made in order to represent their data such that it tells a particular story. Perhaps the best way for students to gain an innate understanding of the human factors underlying data visualizations, and the way that data can be manipulated is, again, to create their own data visualizations. Websites like Our World in Data³⁴ and the Collaborative for Historical Information and Analysis³⁵ provide a wealth of data for students to explore, and they can practice visualization techniques with readily available tools like Google Charts³⁶ or Microsoft Excel.

     Another part of humanizing data is ensuring that students see the human beings and the experiences of these human beings represented in the numbers—that is, students should regularly move between macroscopic and microscopic views of the past. Although data visualizations of various types are useful for viewing large-scale processes and patterns, we risk losing the individual human impact hidden in the aggregate if we do not supplement data visualizations with primary sources like letters, journals, interviews, and photographs. Neil Halloran demonstrates the potential to shift between macroscopic and microscopic views and the powerful story that can be told in doing so through his data-driven documentary, The Fallen of World War II.³⁷ In it, he moves between animated visualizations depicting the number of casualties in World War II and photographic images of the fallen. Teachers can easily humanize data in this way during instruction by supplementing any sustained analysis of a data visualization with primary sources that reveal the thoughts, emotions, and perspectives of individuals represented by the data.

     World history requires a macroscopic lens in order to grasp processes and changes over large swaths of space and time. Data visualizations are the products of such macroscopic views. By allowing students to ignore data visualizations, or by neglecting to teach students how to make sense of them, we hinder their ability to fully access the global past. Teaching data visualizations must be a priority for world history teachers at all levels of schooling. It is important that we turn on the historian's macroscope for each and every one of our students.

Tamara L. Shreiner is Associate Professor of History and Social Studies Education at Grand Valley State University in Allendale, Michigan. Prior to her work at Grand Valley, Shreiner was a world history and big history teacher in Ann Arbor, Michigan. Her research over the last several years has focused on the teaching and learning of data literacy.

¹ See Shawn Graham, Ian Milligan, and Scott Weingart, Exploring Big Historical Data: The Historian's Macroscope (London, UK: Imperial College Press, 2016).

² Graham, Milligan, and Weingart, Exploring Big Historical Data, 1.

³ Douglas Northrop, "Introduction: The Challenge of World History," in A Companion to World History (West Sussex, UK: Wiley-Blackwell, 2012), 1–13; Patrick Manning, Navigating World History: Historians Create a Global Past (New York: Palgrave MacMillan, 2003), 3.

⁴ Graham, Milligan, and Weingart, Exploring Big Historical Data, 1.

⁵ See Tamara L. Shreiner, "Data Literacy for Social Studies: Examining the Role of Data Visualizations in K-12 Textbooks," Theory & Research in Social Education, Vol. 46, no. 2 (2018), 194–231.

⁶ For example, Priti Shah and James Hoeffner, "Review of Graphic Comprehension Research: Implications for Instruction," Educational Psychology Review, Vol. 14 (2002), 47–69; Susan N. Friel, Frances R. Curcio, and George W. Bright, "Making Sense of Graphs: Critical Factors Influencing Comprehension and Instructional Implications," Journal for Research in Mathematics Education, Vol. 32, no. 2 (2001), 124–158; Adam V. Maltese, Joseph A. Harsh, and Dubravka Svetina, "Data Visualization Literacy: Investigating Data Interpretation Along the Novice-expert Continuum," Journal of College Science Teaching, Vol. 45 (2015), 84–90; Priti Shah, Richard E. Mayer, and Mary Hegarty, "Graphs as Aids to Knowledge Comprehension: Signaling Techniques for Guiding the Process of Graph Comprehension," Journal of Educational Psychology, Vol. 91 (1999), 690–702.

⁷ I wish to thank retired AP World History teacher, Ron Schooler, for his helpful comments on this draft.

⁸ See Katy Börner et al., "Investigating Aspects of Data Visualization Literacy Using 20 Information Visualizations and 273 Science Museum Visitors," Information Visualization, Vol. 15 (2016), 198–213; Katy Börner and David E. Polley, Visual Insights: A Practical Guide to Making Sense of Data (Boston: MIT Press, 2014); Cole Nussbaumer Knaflic, Storytelling with Data: A Data Visualization Guide for Business Professionals (Hoboken, NJ: John Wiley & Sons, 2015); David J. Bodenhamer, "Narrating Space and Place," in Deep Maps and Spatial Narratives, ed. David J. Bodenhamer, John Corrigan, and Trevor M. Harris (Bloomington, IN: Indiana University Press, 2015), 7–27; Shreiner, "Data Literacy for Social Studies," 194–231.

⁹ Tamara L. Shreiner and David E. Zwart, "It's Just Different: Identifying Features of Disciplinary Literacy Unique to World History," The History Teacher (forthcoming, May 2020).

¹⁰ Shreiner, "Data Literacy for Social Studies," 218.

¹¹ Ibid.

¹² See James R. Akerman and Robert W. Karrow, eds., Maps: Finding Our Place in the World (Chicago, IL: The University of Chicago Press, 2007).

¹³ See note 9 above.

¹⁴ See J.R. McNeill and William H. McNeill, The Human Web: A Bird's-Eye View of World History (New York: W.W. Norton & Company, 2003).

¹⁵ See Janet L. Abu-Lughod, Before European Hegemony: The World System AD 1250–1350 (New York, NY: Oxford University Press, 1989).

¹⁶ See note 9 above.

¹⁷ Shreiner, "Data Literacy for Social Studies," 219.

¹⁸ See note 9 above.

¹⁹ See for example, Tamara L. Shreiner, "Students' Use of Data Visualizations in Historical Reasoning: A Think-aloud Investigation with Elementary, Middle, and High School Students," The Journal of Social Studies Research, Vol. 43, no. 4 (2019), 389–404.

²⁰ For example, Shreiner, "Students' Use of Data Visualizations in Historical Reasoning," 389–404; Christine Baron, "Using Embedded Visual Coding to Support Contextualization of Historical Texts," American Educational Research Journal, Vol. 53, no. 3 (June 2016), 516–540; J. van Drie and C. van Boxtel, "Historical Reasoning: Towards a Framework for Analyzing Students' Reasoning about the Past," Educational Psychology Research in Social Education, Vol. 20, no. 2 (2008), 87–110; Avishag Reisman and Sam Wineburg, "Teaching the Skill of Contextualizing in History," The Social Studies, Vol. 58, no. 8 (2008), 202–207.

²¹ See Shreiner, "Students' Use of Data Visualizations in Historical Reasoning," 389–404; Peter Seixas and Tom Morton, The Big Six: Historical Thinking Concepts (Toronto, Ontario: Nelson Education, Ltd., 2013); Peter J. Lee and Denis Shemilt, "Is Any Explanation Better Than None?," Teaching History, Vol. 137 (2009), 42–49.

²² For example, Shreiner, "Students' Use of Data Visualizations in Historical Reasoning," 389–404

²³ See note 6 above.

²⁴ Friel, Curcio, and Bright, "Making Sense of Graphs," 124–158; Shah and Hoeffner, "Review of Graphic Comprehension Research," 47–69; Shah, Mayer, and Hegarty, "Graphs as Aids to Knowledge Comprehension," 690–702.

²⁵ David H. Uttal and Kelly J. Sheehan, "The Development of Children's Understanding of Maps and Models: A Prospective Cognition Perspective," Journal of Cognitive Education and Psychology, Vol. 13, no. 2 (2014), 188–200; Jennifer Merriman Bausmith and Gaea Leinhardt, "Middle-school Students' Map Construction: Understanding Complex Spatial Displays," Journal of Geography, Vol. 97 (1998), 93–107; Madeleine Gregg and Gaea Leinhardt, "Mapping Out Geography: An Example of Epistemology and Education," Review of Educational Research, Vol. 64 (1994), 311–361.

²⁶ Shreiner, "Students' Use of Data Visualizations in Historical Reasoning," 389–404.

²⁷ For example, Julianne Coleman, "Elementary Teachers' Instructional Practices Involving Graphical Representations," Journal of Visual Literacy, Vol. 29 (2010), 198–222; Erin M. McTigue and Amanda C. Flowers, "Science Visual Literacy: Learners' Perceptions and Knowledge of Diagrams," The Reading Teacher, Vol. 64 (2011), 578–589; Kathryn L. Roberts et al., "Diagrams, Timelines, and Tables – Oh, My!: Fostering Graphical Literacy," The Reading Teacher Vol. 67 (2013), 12–23.

²⁸ Shreiner, "Data Literacy for Social Studies," 221.

²⁹ Nell K. Duke et al., "Beyond Concepts of Print: Development of Concepts of Graphics in Text, PreK to Grade 3," Research in the Teaching of English, Vol. 48 (November 2013), 175–203; Roberts et al., "Diagrams, Timelines, and Tables," 12–23.

³⁰ Duke et al., "Beyond Concepts of Print," 175–203; Shreiner, "Data Literacy for Social Studies," 194–231.

³¹ See Friel, Curcio, and Bright, "Making Sense of Graphs," 124–158; Bausmith and Leinhardt, "Middle-school Students' Map Construction," 311–361.

³² See Joseph A. Harsh and Adam V. Maltese, "'Seeing' Data Like an Expert: An Eye-Tracking Study Using Graphical Data Representations," CBE—Life Sciences Education, Vol. 18, no. 32 (2019), 1–12; Maltese, Harsh, and Svetina, "Data Visualization Literacy," 84–90. They refer to these steps as "seeing the data," "seeing between the data," and "seeing beyond the data."

³³ For a useful and informative analysis of issues with data that are commonly used in world history, see Dave Eaton, "Poor Numbers," World History through Case Studies: Historical Skills in Practice (London: Bloomsbury Academic, 2019), 229–240.

³⁴ See "Our World in Data," at https://ourworldindata.org/. Accessed October 22, 2019.

³⁵ See "CHIA: Collaborative for Historical Information and Analysis," at http://www.chia.pitt.edu/. Accessed October 22, 2019.

³⁶ See Google Developers, "Google Charts," at https://developers.google.com/chart/. Accessed October 22, 2019.

³⁷ See Halloran, Neil, "The Fallen of World War II," at http://www.fallen.io/ww2/. Accessed October 22, 2019.