Written by Donna M. Bickford, Associate Director of OUR
There is an exciting new initiative at UNC Libraries. Multimodal Librarians “is an initiative at UNC to cross-train the research and instruction staff on digital tools and techniques that can easily be incorporated into undergraduate class assignments. This is a natural extension of the traditional role librarians play in research instruction and it supports the library’s strategic goals of engaging both the entire research lifecycle as well as transformative teaching and learning.”
As part of this initiative, librarians are invited to training sessions to learn how to use digital scholarship tools to support undergraduate research. I had the opportunity recently to attend a session facilitated by Dr. Stewart Varner which focused on using data visualization tools for text analysis. Stewart demonstrated for us Google’s Ngram Viewer, Voyant, and Google’s Fusion Tables.
Ngram pulls information from the Google Books corpus to create its visualizations. You can search in several different languages and you can use words or phrases. I used “undergraduate research” and “inquiry-based learning” for my example.
The Voyant tool creates visualizations based on the text you input. In our session, Stewart uploaded a digitized book from the Project Gutenberg catalog for our demonstration example. Voyant only works with single words, not phrases. You can see the number of unique words in your text. Or, you can use the “keyword in context” tool to see where in your document any particular word appears. There is also a “stop words” function if you want to exclude common words like “the” or “it” — or the title character in the book you’re analyzing. I used the GRC page from our website to generate my word cloud.
Fusion Tables works with data you provide in spreadsheet form. You can also use Google spreadsheets or public data tables. Then you can display the data in several different formats — scatter charts, line or bar graphs, pie charts. etc. I used network analysis format to map the connections between course prefixes and course numbers for our Spring 2015 GRC courses.
While we were experimenting with these tools, Stewart was careful to remind us that at this point we were just generating data points, not making an argument. It’s easy to see, though, how these visualizations might “inspire a research question,” in Stewart’s words. In one project Stewart described, the researcher compared unique words in the texts of two different authors to examine the range of vocabulary one author used in comparison to the other. Stewart also emphasized the need to select the tool that can answer the question you are asking. As a result of the Multimodal Librarians initiative, many of the librarians will be prepared to help students and faculty think through what kinds of research questions and projects will benefit from the use of digital tools.