I decided to write my post about the Twit Lit project, created in 2017. The founder and principal investigator behind this initiative, Christian Howard-Sukhil. She is a digital humanities and postdoctoral fellow at Bucknell University, and has composed a project team of student researchers. Collectively, the team aims to collect, measure, and study the writing community of Twitter. More specifically, their source of data is the amateur tweets, as the project authors believe it is important to record both informal and formal data sources.
The primary research questions of the project were easily identifiable on the project’s website. They have four primary goals, which include the preservation of raw data from Twitter, the creation of visualizations of the data, the analysis of those visualizations, and the development of tools to help other scholars. These research questions/goals all contribute to answering their hypothesis: that collecting amateur data as well as more formal data will close the gap in the literary historical record of society.
In terms of analysis, Twit Lit has created data visualizations that are sourced from text analysis processes scraping Twitter for a total of thirty-five terms. They have cited that they are counting words and completing sentiment analysis, with no evidence of topic modeling through machine learning. From these methods, they have created multiple line graphs measuring the number of writing-related hashtagged tweets over time, tweets using multiple hashtags, and many more.
All in all, I think Twit Lit’s website is professional-looking, clear and concise in their mission and explanation of the project, and transparent in their research. However, I believe their data visualizations are lacking. While they produce numerous line graphs that are informative, I believe they could glean more insights by using other types of graphs, such as scatterplots, bar graphs, bubble charts, and more.
