Fellow(s): Alisa Rod
Mapping, or the visualization of spatial information, is becoming an increasingly effective method for teaching college students to become adept, and critical, users of data (Kim and Bednarz, 2013). In particular, mapping/spatial visualization projects provide students in the social sciences and humanities with opportunities to develop empirical reasoning/quantitative literacy skills (Xie et al., 2018). However, few formal standardized assessment tools exist for the purposes of refining and evaluating various pedagogical approaches, especially with regard to mapping-related assignments and projects within the humanities and social sciences disciplines (Baker et al., 2015). This proposal describes a research project to develop, test, implement, and standardize an assessment tool for the purpose of evaluating the following spatial visualization-related learning goals in a liberal arts context…
Fellow(s): Dorothy Ogdon, Delores Carlito
Quickly and easily creating and evaluating polished and efficient data visualizations is a highly desirable skill for students, faculty, and researchers in educational, professional, and personal settings. Data visualization combines the skills of interpreting visuals (visual literacy), understanding data (data literacy), analyzing the source of the data (media literacy), and understanding the veiled socially-constructed contexts (critical literacy). It involves both creating and interpreting information in visual formats, and it is addressed in ACRL Information Literacy Framework, Mackey and Jacobson’s Metaliteracy, and transliteracy. A knowledge gap in data visualization exists due in part to the idea that techniques and software tools for data analysis and visualization can be quickly and easily self-taught through the use of video tutorials and materials freely available on the internet…
Fellow(s): Ryan Clement
In my current position as Data Services Librarian, I work primarily with undergraduate students from across the social sciences and the humanities. As a liberal arts institution, Middlebury College seeks to give the students exposure to a breadth of subjects in addition to the depth they will experience in their major(s). Like many higher-ed institutions, Middlebury also requires all students to take a first-year course that introduces them to academic writing, reading, and critical thinking. While students who then major in subjects that already focus on working data learn how to produce as well as critically evaluate data visualization, this training is largely centered in courses taken only by students in the major. The training is often highly specialized and less holistic as a result, with students experiencing very different outcomes in their ability to work with visualizations…
Fellow(s): Cass Wilkinson Saldaña
As librarians, how can we create spaces where learners take risks with digital scholarship methods? What would it look like for us to encourage learners to bring their identities and experiences to their work with data, especially in the context of investigating the issues they care about most deeply? What does a distinctly feminist approach to data visualization look like?…
Fellow(s): Sally Gore, Tess Grynoch
Data visualization, as a practice, combines the disciplines of statistical analysis and design. When done well, it enhances our abilities to both better understand and summarize a particular data set, as well as communicate the same to an audience. Without statistical analysis, data visualization is “an exercise only in illustration and aesthetics,” while without good design, it fails to inspire (Yau, 2013). It is the intersection of these two disciplines that creates the interesting question of whether or not data visualizations can – or should – maintain neutrality in the messages they convey, in particular messages that evoke an empathetic response…
Fellow(s): Negeen Aghassibake
I am a data visualization librarian working in a large health sciences library in the middle of a large city in the United States. I am also a brown woman working in a profession that is predominantly (at my most recent check, nearly 88%) white. More specifically, I engage in work that is dominated by the tech industry, which also lacks commitment to diversity, equity, and inclusion. I say this not parrot facts and statistics about the library and technology professions, but rather to provide context to my interest in the topic of critical data visualization, particularly in consideration of race, gender, sexuality, and ability status. “Best practices” and tutorials for data visualization often follow the advice of white, Western, sexist, cisgender, heteronormative, and ableist lines of thinking. Libraries have an opportunity to step into the world of data visualization and empower historically marginalized and underrepresented voices in big data. I would like to explore the intersections of social justice issues and data visualization. If we’re visualizing the future, we need to consider who we’re visualizing for, what we’re visualizing, and why we’re visualizing…
Fellow(s): Amy Sonnichsen
For my project, I am interested in researching and developing a set of criteria and/or guidelines for the creation of data visualization in relation to community health data. The focus of these guidelines would be to establish methods for creating community health data visualization and to communicate the nuance and ethical implications of revealing disparities among diverse communities. The purpose of the guidelines would be to establish the responsibility a researcher takes on as a visual storyteller, especially concerning the decisions made when manipulating health data. My further intent in creating these guidelines, would be to engage students’ critical analyses in asking the appropriate questions of the data, and then translating those questions into meaningful visualizations…
Fellow(s): David Christensen
Many organizations engaged in social justice work strive to allocate resources for their communities equitably. Increasingly these organizations seek to include visual and non-visual representations of inequity data in decision-making processes—as use of data replaces use of anecdote. One approach to representing inequity is through opportunity index mapping, whereby several demographic data points are aggregated into a single index score by census tract. The index scores of multiple census tracts are then (often) combined to conform to the geography of existing organizational infrastructure, such as the service area of a public library branch, or boundary of a public school. The aggregation of multiple data points and the translation of multiple geographies into an ‘existing organizational geography’ may have serious implications for data visualization. For example, if you combine a ‘high opportunity’ geography with an adjacent ‘low opportunity’ geography in a visualization, you might distort the way both geographies are treated in decision making. If practical opportunity mapping is to be successful, it must avoid this sort of distortion and still be visually understandable to decision-makers. Which data inputs are selected, and how any existing organizational geographies are visualized may have real-world implications for how organizational resources are allocated.
Fellow(s): Megan Ozeran
For my individual project, I plan to investigate and define the ethics of data visualization. These ethics have not been formally defined, and it is crucial that we do so. Current technologies make data collection, data analysis, and data visualization easier than ever before, even as the amount and variety of data continue to expand. Computer-mediated data analysis is no longer simply in the domain of computer scientists, as more tools and software are made available to researchers with a variety of technical expertise…
Fellow(s): Jo Klein
Most data visualization processes are made up of four basic and often iterative steps: 1. Data and an appropriate visualization tool or method must be found and obtained, 2. Data must be processed and reorganized to fit the tool, 3. The visualization tool must be used to make a product, and 4. The product must be distributed and made accessible. Metadata can be leveraged to help at any step in this process, from finding appropriate data and guidance for a tool using keyword searches to making a visualization findable by others once published…
Fellow(s): Matthew Sisk
Representing uncertainty is difficult to do in many data visualizations. This is particularly true in spatial data visualizations and cartography, where at least two of the dimensions (x and y) are typically unavailable for representing uncertainty. Despite this, most spatial data have some measure of uncertainty associated with them. This can be spatial uncertainty (e.g. gps error, ambiguity in the border between polygons or raster resampling effects) as well as data uncertainty (e.g. error terms in modeled data or uncertainty measures that come from survey data). The available literature on uncertainty in spatial visualizations tends to be highly complex and focused on advanced, highly computational techniques, making it a daunting topic for novice or intermediate users…