Fear not statistics!

I have never been good with numbers and statistics. I actually dreaded them for a long time, because anytime I came across numbers or statistics, I felt that there was something obvious I should understand, a pattern I should see, but I just did not. During my studies, I tried to stay away from statistics courses, and instead chose courses were I could engage with reality through words. Little did I know that I would have to work with statistics when I started my Master's degree in... French Literature!

I became a research assistant for a retired professor at the University of Ottawa when I started my Master studies in French Literature there in 2012. I had not idea about what type of work I would actually be doing, but when I heard that a professor was looking for someone to work on 18th century manuscripts recently acquired by the University Archives, I went to meet her and basically asked: where do I sign?

The professor I was working for was a very busy researcher and gave me very few directions as to what I should be doing with the manuscripts, except that they were constituted of many disparate texts and that we should be trying to find common threads and see how they all fit together.

As a first task, she asked me to inventory all the people that the author named in the manuscripts (she was a member of many scientific and literary circles and let me tell you, she knew A LOT of people). The manuscripts were about 2000 pages in total, so I spend the next few weeks (read months) reading and noting all names and their occurrences in an Excel spreadsheet (see below part of the spreadsheet).



After I was done, I sent my spreadsheet to the professor, still unsure about why I spent some much time gathering this information. When I met with her, she was really excited and that the spreadsheet would be perfect for a communication at the upcoming Canadian Society for Eighteenth Century Studies annual conference.. I was stunned. AGAIN there seemed to be, in that mass of data, an obvious pattern or truth that she was seeing and I was not... After discussing with her, and actually spending some time look beyond the names and numbers, I realized that it was actually true: there were many things that could be said and understood from all this statistical data about the manuscripts, and even about the life of the author. Moreover, there a whole dimension of the work we could not have understood without statistics.

I ended up presenting at the conference, and my presentation was actually titled "The surprises of an index". Yes, you read that well. And to my pleasant surprise, I actually had little work to do in order to convince other researchers at the conference that such statistics enabled us, for example, to qualify the manuscripts as autobiographical (this is simply stated here, but one our argument was that members of her family were cited the most often and used to tell her personal story) and to map a network of her personal relationships, but also a larger network of relationships between people of the scientific, literary and political circles and the high society of Paris during the 18th century.

This experience absolutely changed my views on the use of statistical data in any type of research. And even if patterns and truths remain less obvious to me when hidden in statistics, I still say to myself: "Fear not statistics"!