Data Analytics in Historical Sources

Monday, August 28, 2017 - 5:00pm
Jan Hoffmann

This blog post is written by Jan Hoffmann, a student at Manhattan College who was part of one of our NYG&B Labs teams this summer. The team assisted Professor Adam Arenson in his project tracking and visualizing migration patterns of African North Americans prior to, during and after the American Civil War. A portion of the project focused on locating Canadian-born African North Americans in the U.S. Federal Censuses for 1860, 1870 and 1880. The project is still a work in progress, but we are excited to have the students share their updates in a series of blog posts over the next month.  


As a Manhattan College student majoring in Management and Business Analytics, my summer research with Dr. Adam Arenson, Associate Professor of History, and Dr. Musa Jafar, Associate Professor of Accounting and Computer Information Systems has been on an interesting topic in a completely different area than I was used to.

My goal was to visualize locations, migration routes, and social networks among African North Americans crossing the U.S.-Canada border during and after the American Civil War using RStudio, Excel, and Tableau.


Processing the historical data

I pulled the geolocations of our data frame by using Google Maps API to get the latitude and longitude information of the data. This was the perfect task for me to apply my knowledge from this year’s “Data Mining” course taught by Prof. Jafar, who also assisted and helped me throughout the project.

Given the irregularities in historical data and the need to create a new way to visualize this data, many technical problems occurred that could not have been learned beforehand. We could only find and solve them by doing the work—this is why this research project was such a good experience for me.


Visualizing the data

Once I got the longitude and latitude information, I saved the files in a CSV file before importing them into Tableau to visualize.

In Tableau, I mapped the Canadian-born individuals we found in the U.S. census, then sorted by census year and counted them by region. This gave us very interesting insights where African North Americans went after crossing the U.S.-Canada border.

This map shows regional clusters of Canadian-born African North Americans in the Great Lakes region, recorded in the 1860, 1870 and 1880 censuses. 

This map shows regional clusters of Canadian-born African North Americans in the Northeast, recorded in the 1860, 1870 and 1880 censuses.

Our data frame showed significantly more Canadian-born African Americans moving to California in the 1860 census than in the 1870 census, a result worth exploring more:

Throughout the maps, trends can be seen of people moving to urban areas rather than rural ones, which may reflect the Reconstruction years or this population in particular.

I will leave the interpretation to the experts of this field and I will continue my work focusing on the following census years, tracking these migrations and finding statistically significant differences.


Are you interested in learning more about African-American genealogy? NYG&B members can watch a replay of an amazing talk from Kenyatta Berry that happened at the NYG&B in November 2016. Her presentation, From Virginia to Upstate New York: An African-American Family's Journey to Upstate New York offers another look at a subject closely related to what our Labs team is researching. Our New York Knowledge Base also contains a bibliography for African-American research, and a recent blog article listed some essential resources for African-American research in the NYG&B Record.

By submitting this form, you accept the Mollom privacy policy.
Wonderful research. I have done considerable research in census records, 1850-80, for African Americans living in NYS. If they listed their birthplace as a southern state, Canada, or unknown, I consider that a red flag for likely status as a refugee from slavery. In some cases, we have found the same person who lists his/her birthplace as a southern state in one census and New York State in the next. In one case in Cayuga County, Jerome Griger listed his birthplace as unknown (1850), West Indies (1855), Spain (1865), and New York (1870). It seems reasonable to conclude that place of birth as listed in census records for freedom seekers was as much a political statement as it was a statement of actual fact. Love your research, Jan!