Data visualization

Overview

The presentation of data, be it statistics, a timeline, narrative structure, or other form of argument, is often accompanied with a visual presentation. This is called data visualization. In data visualization, the visual element of a presentation has been designed to ensure that all parties in a discussion or decision-making process understand and can correctly interpret the information being presented to them. For some learners, data visualization is the best way to understand information, whereas for others, data visualization simply helps to contextualize facts that are presented orally or in written form.

Communication scholars are interested in data visualization for many reasons, ranging from the ways that data visualization affects persuasion, the ethics of data visualization, and ways in which data visualization can be best incorporated into visual and written presentations. For example, Bevington-Attardi and Ratcliffe have studied the ways that United States census data is commonly displayed through maps and software that processes large data sets. These scholars frequently focus on the ways that descriptive statistics, those statistics that summarize information about a group, can be presented in a visual form. Scholars are also interested in the ways that more traditional data sets, such as a national census, are presented using tables and charts. There are many ways that census data, such as the age of residents in a region could be reported, ranging from bar graphs to tables to pie charts. Communications scholars are interested in reasons why a speaker or presenter would choose one presentation style over another. The need to make a critical and correct decision about how to display information is illuminated in the following example.

Imagine a researcher needs choose to present the different ages in a community through a bar graph or pie chart. Imagine a community where 20 percent of the population is between the ages of 0 and 18, 20 percent between the ages of 30 and 40, and 60 percent over the age of 60. A bar chart might show significant gaps in the population—such as that there are not any residents between 19 and 30 in the community. However, a pie chart, would divide the circle into three groups, this would show a significantly larger population of 60+ year old residents, but would not highlight the lack of residents between 19 and 30. Both data visualizations show the same information, but the take away for the audience is different. In the bar chart, the community is presented as obviously missing a significant demographic. Whereas in the pie chart, the community is presented as a whole circle, which has a high number of 60+ residents, but is not necessarily missing any demographic groups. The decision regarding which graphic to use is critical to public policy makers who need the public to decide on new policies.

Theoretically each member of the public could analyze the census data alone, and then consider a set of proposals. However, few members of the public have the time or desire to do so, which means that using data visualization is a necessary way to present complex information in an easily understandable manner to enable better decision-making. These tools are used in the portrayal of healthcare information as well as educational tool kits. Among the many advantages of data visualization is its ability to highlight change, such as an increase in residents of a demographic group or the way that a child has improved in an academic course. Aaron Marcus has found that even on an individual level, data visualization can be used to encourage patients at risk of obesity to make healthy lifestyle decisions and changes.

Advances in computer science have allowed scholars and researchers to broaden their study of information, communication networks, and the way that data visualization is used to explain our world. Oftentimes computers are used to gather data sets scholars call "big data." Sets of big data are increasingly easy to produce as computers are able to gather complex sets of information from many sources. Researchers have difficulty sorting through and then presenting that data without the assistance of computer software. For example, when a computer is used to collect information regarding Internet and satellite communication, it can be hard to verbally explain the network of communication between all different users; however, that information can easily be displayed using a visual description of the communication network. This can be a helpful tool in understanding the ways that online platforms and forums, such as Reddit, are used by various communities and audiences. Jessica Peter's study of subreddits, Reddit's single topic forums, demonstrates how online communities are built, the types of participation by commenters on Reddit, and the interconnection between various topics and communities.

Some scholars are interested in the ways that data visualization is perceived by audiences, while others are interested in the ways that presenters and writers construct data visualization tools. The development of software, online tools and platforms, and presentation technologies all have pressed communication scholars to continue developing their theories and ideas regarding data visualization and its effects on society. Additionally, basic spreadsheet software programs, such as Microsoft's Excel make it easy for many presenters to produce their own graphics. This ease of production saves companies and presenters a good deal of time and money that was once spent having graphic designers and other professionals produce data visualization tools. However, the ease of access to programs such as Excel also puts pressure on presenters to ensure that they have chosen the best data visualization tool from a range of options.

Further Insights

Data visualization is an excellent tool for encouraging audiences that are new to a topic to quickly understand the key components and challenges faced by policymakers, advocates, educators, and other concerned members of the public. For example, Rall et al.'s study examined human rights advocates' use of data visualization to help the public understand where and when human rights violations occur. While many human rights advocates agree that data visualization is a critical tool to engage supporters, they disagree on which tools are best to use to make their points clear. One point of contention is the turn from narrative arguments to those that are supported by statistical data. This turn has occurred in part because it is now easier to collect statistical data and find ways to present it to the public. Some advocates believe that using big data, which is presented through data visualization, allows supporters and the public to quickly grasp the magnitude of a human rights violation. However, others worry that the human qualities and uniqueness of each human rights violation is lost when they are presented in a summarized statistical format. For this reason, scholars are interested in working with human rights advocates to find a middle ground, which preserves the specificity of human rights abuses while also capturing the magnitude of those abuses through data visualization.

Communications scholars are also interested in the ways that the public is expected to interpret and understand visually presented data. For example, Grossman analyzed the ways that weather data is presented to the public. While the public once received information about the weather in a very condensed form, it is now possible for weather reporters to display satellite images of oncoming storms and detailed information about the probability of a storm affecting a community. While this information may be helpful for some, it is overwhelming or misinterpreted by others, possibility leading to poor decision-making. For this reason, scholars such as Gharesifard, When, and van der Zaag studied the ways that citizens collect and understand weather data so that they can enable more user-friendly and accurate representations of future storms and climatic events. In this way, communications scholars have assisted weather reporters in editing the information they provide to the public. This process of editing is difficult because experts want to provide as much information as possible, yet they must keep in mind that their audience members are not experts. The goal of data visualization is to make sure that the audience can appreciate the information that they have received and can act accordingly when they receive that information. For example, if a storm is on the way, it is most critical that the weather reporter quickly provide information about when the storm will arrive and what precautions are necessary. The reporter must be careful not to become too excited or distracted by the ways that this information can be displayed and instead must focus on the most efficient and easy to understand display.

Issues

Communications scholars are also interested in the ways in which data visualization can and has achieved poor results. For example, they are concerned with the ways that data visualization can be used in ways that unintentionally cause audiences to make poor decisions. One of the most commonly cited examples is Tufte's examination of the Challenger space shuttle explosion in 1986. Tufte argues that a presentation made days before Challenger's lift-off included information about a risk of explosion. However, the information contained in that presentation was presented in a confusing way, which prevented key decision makers from understanding the risk that was involved in Challenger's launch. Tufte's examination of the Challenger explosion has set a standard among communications scholars who are interested in ensuring that information is presented in the most ethical and understandable way possible. The Challenger example demonstrates what can go wrong when information is presented in a confusing way, even when no one was trying to misinform decision makers.

Using Tufte's analysis of the Challenger disaster, along with his many other studies of data visualization, scholars have designed complex studies, computer programs, and other methods of designing new data visualization and ensuring that data visualization correctly presents data. For example, Finch and Flenner have examined the way that data visualization can be used to examine a library or archive. After producing a data visualization tool that displays an archive, librarians might be able to spot holes in the collection or find areas where there are not enough or are too many books on a particular topic. Or, they might code books in different colors to chart how frequently they are checked out by library patrons. From this data visualization, librarians would be able to produce a more complete library collection, and through that collection better serve their patrons.

A library may seem like trivial place to focus so many data visualization tools, but think about the types of resources housed in a library. A hospital needs to ensure that doctors have access to information regarding many different types of diseases. One way to ensure that they have good coverage might be to code articles and resources in different colors or as part of a pie chart, which could be used to represent a whole body. From this depiction, librarians could tell if there was an overrepresentation of one field of medicine or a lack of information regarding a problem doctors are frequently asked to diagnose. Similarly, factories may want to inventory machinery, or restaurants may want to account for the different types of vegetables that they have on hand. While all of this information could be presented verbally or through a set of statistics, using data visualization is a strong way to ensure that an audience is paying attention, can understand the information provided, and can act in a way that utilizes that information to its fullest potential.

Data visualization became critical during the COVID-19 pandemic, which began in the United States in 2020. Data was used to convey information such as the daily totals of confirmed cases, the global reports of cases and deaths, the hospitalization rates for various age groups, and how to properly wear a mask. The many types of data visualization used during the pandemic included visualizations embedded in text, infographics, and dashboards. Of these, dashboards were especially helpful because they allowed users to get a glimpse of important information and also access more complex research.

Bibliography

Bevington-Attardi, D., & Ratcliffe, M. (2015). Data visualization at the US census bureau–an American tradition. Cartography and Geographic Information Science, 42(sup1), 63–69.

Crisan, A. (2022, June). The importance of data visualization in combating a pandemic. American Journal of Public Health, ajph.aphapublications.org/doi/full/10.2105/AJPH.2022.306857

Finch, J. L., & Flenner, A. R. (2017). Using data visualization to examine an academic library collection. College & Research Libraries, 77(6).

Gharesifard, M., Wehn, U., & van der Zaag, P. (2017). Towards benchmarking citizen observatories: Features and functioning of online amateur weather networks. Journal of environmental management, 193, 381–393.

Grossman, S. J. (2016). Ugly data in the age of weather satellites. American Literature, 88(4), 815–837.

Marcus, A. (2015). The health machine: Combining information design/visualization with persuasion design to change people's nutrition and exercise behavior. In Mobile Persuasion Design (pp. 35–77). Springer, London.

Midway, Stephen R. (2020 Dec.). Principles of effective data visualization. Patterns, 1 (9), doi.org/10.1016/j.patter.2020.100141

Otten, J. J., Cheng, K., & Drewnowski, A. (2015). Infographics and public policy: Using data visualization to convey complex information. Health Affairs, 34(11), 1901–1907.

Peter, J. (2015). Visualizing Reddit: Exposing user communication patterns through data visualization on Reddit.com. Blucher Design Proceedings, 2(2), 1668–1674.

Prakash, T. G. (2016). Data visualization and communication by big data. International Research Journal of Engineering and Technology (IRJET), 2(2).

Rall, K., Satterthwaite, M. L., Pandey, A. V., Emerson, J., Boy, J., Nov, O., & Bertini, E. (2016). Data visualization for human rights advocacy. Journal of Human Rights Practice, 8(2), 171–197.

South, E. & Rodgers, M. (2023, Aug. 17). Data visualisation in scoping reviews and evidence maps on health topics: A cross-sectional analysis. Systemic Reviews, 142, doi.org/10.1186/s13643-023-02309-y

Williamson, B. (2016). Digital education governance: Data visualization, predictive analytics, and "real-time" policy instruments. Journal of Education Policy, 31(2), 123–141.