On May 29, the discovery of the remains of 215 children buried at the site of a former Indigenous residential school in Kamloops, British Columbia, created a shock wave across Canada. However, the phenomenon had previously been thoroughly documented : the Truth and Reconciliation Commission of Canada issued a report in 2015 revealing that at least 3000 children that attended residential schools had died – and the cause of death of half of them remains unknown, due to lack of documentation. Our project shed light on what we know – and still don’t know – about those children.
While the information in this project wasn’t new, it was largely unknown, and that is why we felt it was important to bring it to light. After the Kamloops discovery, many people were starting to realize this facet of Canada’s past. Six days after this project was published (June 18th), another discovery of anonymous graves occurred in Saskatchewan, once again bringing the matter up to date.
Radio-Canada Toronto (CBC French Services) contacted us to discuss the project on their radio show (see project link 4, 8:50). It was the first time that a radio station invited us to speak about a data visualization project on air. Since we are located in Montreal and the radio show was in Toronto, it made us realise our project had an impact in other provinces.
Since we are located in the province of Quebec, our project shed light on former residential schools that were located in our province, making Quebeckers aware that these institutions also existed near them, and that it is not only something that happened in Western Canada, where almost all the anonymous graves have been discovered to this day. We described how many schools were located in Quebec, where they were located and for how many years they were in function.
The project was also the main subject (front page) of our newspaper’s Saturday’s special paper edition « Perspectives » (see project link 3, page 13-14-15 of PDF).
Olivia, the data visualization developer, mostly used D3 to create the project’s visual. For my part, the majority of the data was extracted by hand. I first tested some charts in Google Sheets ahead of time, to test their potential and test which representations read best for our readers. Once the plan is established, it reproduces the selected graphics with D3. We also pulled a map and a graphic from a geojson file from the Canadian Encyclopedia.
What was the hardest part of this project?
The hardest part of this project was certainly the access to the data and the manual labor that it involved. All the data can be found in the TRC report, entitled “Missing Children and Unmarked Burials” (see project link 5). It is a PDF document filled with tables and charts. Most of the data has been taken from the chapter on statistical analyzes. Several data were drawn by hand from the tables, and sometimes from the text.
A lot of the raw data displayed in graphs were missing from the document – not even as an appendix to the document. Some charts like histograms and bar charts happened to be very relevant, but it was impossible to reproduce them or deduce the data since the axis was not always the most precise.
After much research, we thought about using the very same graphics from the PDF document (screenshots with permission of the authors), but we concluded that it would affect the visual aspect of the project too much. I finally took steps to get in touch with the author of the statistical chapter to request the exhaustive data for the report. The process was tedious, as the commission’s authorization was required as well. I finally got all the data for the report after a few weeks.
What can others learn from this project?
I get the impression that when people think of data journalism, they immediately think of big databases, open data, or huge files that need to be cleaned up. I think data journalism is also about going into reports and paper documents to look for a little gold mine. Especially to cover old historical events, many archives contain a wealth of relevant data to put into context. I believe it is important to consider it, even if it implies the data would be in textual form or in PDF formats, and that it takes more manual work and time to convert.
I also think that journalists can learn that it is relevant to illustrate data that is missing: our reflex is often to cover and illustrate tangible data, but less often to visualize unknown or missing information. In such a historical context, choosing such an angle demonstrates a very different side of history, and honors those who are often forgotten by statistics.