Religions in the College News

Motivations and Process

This graph shows how frequently different religions were mentioned in the College News corpus. Protestant Christianity was excluded because Bryn Mawr College was founded as a Protestant Christian (specifically, Quaker) institution and this project aimed to bring attention onto other religious groups.

The motivation for this project was a desire to know more about religious diversity in the history of Bryn Mawr. While I quickly realized that I would not be able to get a huge amount of information about different religious practices at Bryn Mawr in the early 20th century, I was still able to explore attitudes toward religion and ways that students were engaging with religion on campus and through the College News. One of the things that stuck out to me was the change in usage of the term “interfaith.” It was hardly used at all until the late ’50s, when it started being used very frequently. I learned during my research for this project that the rise in the usage of this term coincides with the formation of a student “Interfaith Association”, which invited speakers to talk about different religions, among other things. The activities of the Interfaith Association were discussed often in the College News, hence the frequent appearance of the term.

The data for this visualization was acquired through a combination of keyword searching and topic modeling. I identified potential keywords by uploading the corpus into Voyant. I generated a list of religions to search for using lists for Wikipedia. Ultimately, my list most closely reflected this list of religious groups by proportions, but I also referenced a more detailed list of religions, using the categories within to expand my search. I used the terms from these lists to search in Voyant’s frequency visualization tool, noting each search that was getting results. I performed a keyword search to identify instances where religions were mentioned in the corpus and extracted the 250 characters surrounding each keyword as context.

I used the Topic Modeling Tool to generate topic data for each context chunk and added the most relevant topic for that instance to my dataset. Then, I aggregated data by year and identified the most frequent topic for each religion in each year. Finally, I used Altair to display the data. I modeled my chart after the “Natural Disasters” visualization from the Altair example gallery and added tooltips to show the specific numbers as well as the topic data for each point. I exported the chart as a JSON object and then embedded it in the site using Vega.

Topics and Associated Keywords