“Hashtag History” is presented as a digital dissertation for three reasons:
- The use of digital tools to collect and analyze data enabled a more extensive case study than is typical for historical thinking research.
- The digital format preserves dynamic content, such as tweets, videos, and GIFs, in their original form.
- Offering the dissertation publicly and online ensures transparency and accessibility. The students who produced the dataset can access how I used their data in the final form of the project. “Hashtag History” also conforms to web accessibility guidelines to highlight the need for greater attention to accessibility in both traditional and digital dissertations.
Digital Tools Forward Historical Thinking Research
“Hashtag History” contributes to the scholarship of historical thinking by employing data from a larger pool of participants than most historical thinking research. Where the majority of studies focus on 15 to 50 students, this project’s dataset includes the work of 150 students. Smaller studies are advantageous for discerning individual students’ progress, but make it difficult to see how students socially and collectively construct their understandings of history. Examining data produced by a larger pool of participants illuminates broader trends in how students as a whole interacted with history and each other.
This dissertation is not the only larger-scale study of historical thinking. However, my status as an individual researcher instead of a member of a research team distinguishes this study from more substantial research like that of the CHATA project (320 students) and research led by Adele Nye in Australia (1,455 students). Independently studying 150 students across a semester would have been exceptionally challenging without the use of digital tools. Employing web-scraping to harvest data and R packages to clean, analyze, and visualize data allowed me to collect and examine a dataset from a larger set of students than I might have done through traditional interviews or questionnaires. While I utilize close reading throughout this dissertation, I primarily processed the 11,454 collected tweets and 74 blog posts using digital methods, such as web scraping and text mining.
For the most part, text mining using sentiment analysis and word frequency studies confirmed patterns documented in previous historical thinking studies, such as the impact of emotion in the classroom and students’ preference for content with clear personal relevance. But focusing on a dataset culled from the web and analyzed with digital methods also offered new opportunities. Most notably, web-scraping allowed me to collect both text and media. Using the harvested media, I was able to track students’ use of visual media and the impact media has on their emotions, attention, and historical skills. Viewing visual media as a sociocultural influence offers a new angle for historical thinking research.
Understanding how large groups of students collectively express emotions, channel their attention, and adapt visual media is useful for considering the goals of survey classes. For many students, a survey course comprised of a hundred or more students will be their only encounter with history as an undergraduate. Given the size of general education history courses and the shrinking number of history majors, it is essential for researchers to consider how students process history together in survey courses. Historical thinking is beneficial beyond the major, but in order to communicate that to students (and parents and institutions), history educators need to understand how best to engage and develop students’ approaches to history on a broader scale. Research on a scale similar to the size of undergraduate survey courses offers valuable insights to instructors of these courses.
Preserving Dynamic Content
In addition to collecting and analyzing data using digital tools and methods, “Hashtag History” takes the digital form of a website. As noted above, the dataset that forms the primary source material for this dissertation contains tweets, blog posts, and visual media. By presenting the project as a website instead of a static document, the data can be presented in dynamic forms. This offers users opportunities to interact with the media and datasets. Tweets and threads are clickable, datasets can be downloaded and transformed by other users, and the code behind visualizations is easily shared. Presenting “Hashtag History” publicly and online ensures that the primary source materials remain true to the original form and are fully accessible to readers.
In addition, much of the dynamic content included in the dissertation has been preserved using the Internet Archive’s Wayback Machine. Articles and e-books are not archived, as these are frequently paywalled and subject to institutional oversight. I have worked to prevent the loss of open-access, web-based media, however, by creating stable links in the Internet Archive’s database. Archived links for content outside the dataset, such as web pages and videos, are included in footnotes. Archived links for students’ blog posts are available in the Blog Post Master List.
Students’ tweets, though, are not preserved in the same manner. While the basic content of tweets is preserved in the All Tweets Dataset, the tweets and accompanying media have not been archived out of respect for students’ online agency. All data was public at the time of collection (and therefore fair game for inclusion in this project), but students were under no obligation to maintain their Twitter accounts. It was and will remain their choice to allow this piece of their digital identity to persist. As a result, some tweets included in the dissertation have already been protected or deleted; I suspect more will disappear in the future. When they do, I will replace embedded media with the static text of tweets.
Student Access & Web Accessibility
Accessibility is a core value for this project. The 150 undergraduate students enrolled in my world history course in Spring 2017 produced the data for this project through their completion of course activities. Out of respect for their willingness to share their work with me, I share my work in return. Presenting an open, web-based dissertation ensures that students can review the ways I have used and interpreted their data. The students may never read the dissertation in full, but it is crucial that they could access the project if they wished.
The project also models web accessibility. This digital dissertation adheres to W3C’s POUR principles, meaning the site aims to be Perceivable, Operable, Understandable, and Robust for as many users as possible. To that end, the website utilizes an accessibility toolbar, an accessibility-ready theme from WordPress, alternative text for images, accessible tables for all visualizations, ARIA attribute footnotes, and plain, clear language as often as possible. Inclusive design remains rare in scholarly work and teaching, despite opportunities for scholars to build their skills independently (using sites like WebAim) or collectively (through workshops and training institutes like the Digital Humanities Summer Institute). My hope is that calling attention to web accessibility through this dissertation will encourage greater attention to accessibility in both traditional and digital dissertations.