Project description

I work as a data reporter for CBC/Radio-Canada. There are two official languages in Canada: English and French. Because I am based in Montreal, Quebec, I work mostly in French, but I often translate my projects.

For my portfolio, I choose six completely different stories.

\”Université: payer pour avoir moins\” is about the steep increase of special fees in Quebec universities. When I asked the government for the financial data of all of Quebec\’s universities, my request was declined. But I found a source who accepted to give me 20 years of data. I had to understand the accounting data, cross-reference it with the number of students and professors in each university, for each year, to be able to do an analysis. Then I needed to code my visualization to explain my findings in a simple and clear way to our audience.

\”How the CAQ won the Quebec provincial election\” was published after our last general election. I had very detailed data on who voted for which party and when. I created a simple and efficient data to explain how the Coalition Avenir Québec succeeded to win the election, even if just a quarter of the voters chose them for their policies. On the same template, I published \”Et si nous avions eu un mode de scrutin proportionnel?\”, which showed of the CAQ party won the election, even if most of the voters didn\’t vote for it.

The goal of \”On a cloné les chefs\” was to give the possibility for the voters, during our last general election Quebec, to ask questions directly to the political leaders. I asked 50 questions on 9 different topics to each leader. Then their responses were loaded into four chatbots.

\”Et si vous aviez investi dans le cannabis\” is about the legalization of marijuana in Canada. I heard so many people saying that they should have invested in marijuana producing company that I decided to create a stock market simulation for them, with a vertical timeline. It brings them back in 2014, with $100 to invest.

And \”Forget the snowy winters of your childhood\” is a story on climate change. I scraped the data from Environment Canada\’s weather stations. And I produced an analysis of how winters changed through the decades because of global warming. I also coded a series of visualizations to explain it.

What makes this project innovative?

Each project has its own challenges. For "Université: payer pour avoir moins", I had to understand 200,000 rows of accounting data. The story "Forget the snowy winters of your childhood" started with a 16 million rows dataset and I had to calculate everything in decades to compose with the high variability of weather data. The analysis is always a challenge. I had to improve my Python skills a lot to be able to do precise calculations, in a reproducible way, into Jupyter Notebooks. For the visualizations, the biggest challenge is always to create an experience that works perfectly on mobile, since a huge proportion of our readers uses this kind of device. Having narrow vertical screens and less powerful processors is a tough deal. I perfected my JavaScript and learned how to use canvas elements instead of svg elements, to improve performance. I also think my dataviz with a vertical screen in mind, generally for iPhones 5, just to be sure that it'll fit everywhere. The vertical timeline of "Et si vous aviez investi dans le cannabis" is a good illustration of this.

What was the impact of your project? How did you measure it?

The story about the increase of special fees in university had a big impact. The minister of education had to go live on television on the evening of publication day to justify her work. "How the CAQ won the Quebec provincial election" and "Et si nous avions eu un mode de scrutin proportionnel?" were also huge hits. Even if we are a small province with 8 million people, around half a million page views with those two stories. And "Forget the snowy winters of your childhood" is so far one of the most read story of 2019 on our website. In general, we see that people tend to stay longer on our data visualization work than on our other stories. This is especially true on mobile devices since we work a lot on having a great experience for small screens and less powerful processors.

Source and methodology

The data for "Université: payer pour avoir moins" came from a professor whom I convinced to give me his dataset. The government refused to give me the information. "How the CAQ won the Quebec provincial election" and "Et si nous avions eu un mode de scrutin proportionnel?" were based on electoral data and data provided by VoxPop Labs. It's a non-profit company, specialized in big surveys (hundreds of thousands of people) during electoral campaigns. For the story titled "On a cloné les chefs", I am the source for the data. I asked 50 questions on 9 different topics to each political leader, during the Quebec general election. "Et si vous aviez investi dans le cannabis" is based on Yahoo Finance data. The data for "Forget the snowy winters of your childhood" comes from the Environment Canada website. Concerning the methodology, when the story is quite complex, I write research hypotheses and I write into a Google Doc how I plan to find whether they are true or false, with data. Then I share this document with researchers and professors who are experts on the topic. It's a great way to avoid mistakes and to be sure that my analysis is solid. Usually, some of the experts also end up commenting on my results in my published stories. That's what I did for "Université: payer plus pour avoir moins?" and "Forget the snowy winters of your childhood".

Technologies Used

I code my web scrapers in Python or Node. The analyses are done with Python, into Jupyter notebooks. The templates for all of my projects are my own, build on with a custom version of Parcel, specifically for my needs. The data visualizations are made with D3. When the calculations are not too intensive, I animate SVG elements. Otherwise, I draw into canvas elements to improve performance.

Project members

Melanie Julien, Eric Larouche, Santiago Salcido, Francis Lamontage, Andre Guimaraes, Marc Lajoie.


Additional links


Click Follow to keep up with the evolution of this project:
you will receive a notification anytime the project leader updates the project page.