When he was doing his master’s degree in astrophysics, Philippe J. Fournier spent entire nights at the Mont Mégantic observatory.
During the day, he slept. When the sun set, he would cool his telescope with liquid nitrogen to map small corners of the cosmos. His hunt: the Wolf-Rayet stars, the largest and most unstable in the universe, destined to become black holes. He spent nearly 200 hours there, alone with a technician, converting points in the sky into data.
Listening to Philippe J. Fournier look back on this period of his life, we understand the common thread in the journey of this CEGEP professor of physics who became an expert in survey analysis.
How do we move from stellar corpses to the subtleties of voting in Sainte-Foy? By becoming passionate about data and figures. In both cases, he pieces together information to reconstruct an image of reality.
“At the observatory, each speck of light delivered a small electric shock that generated data, 0s and 1s. Together, they created a 3D matrix. What I do with my Qc125 site is not that different…”
And he does it very well. In the last Quebec elections, he predicted the results of 118 of the 125 ridings. He only made two mistakes. “For the other five, I assessed that it was a pivotal constituency, with an equal probability of victory,” he specifies.
He is not interested in politics to advance ideas. Its concern is that of statistical rigor. This rigor and this talent for popularization make him today an indispensable analyst of Quebec and Canadian politics.
Philippe J. Fournier grew up on the South Shore. When he was 9 years old, his family moved to Saint-Augustin, a suburb of Quebec.
PHOTO ROBERT SKINNER, THE PRESS
Philippe J. Fournier
Gifted in science, he enrolled in physics at McGill University. He continued with a master’s degree in astrophysics. “Beyond the fascination that the sky exerts, from a scientific point of view, it is also a beautiful synthetic discipline. It integrates all fields of physics, such as optics, relativity or electromagnetism,” he explains.
After his graduate diploma, he obtained a part-time teaching position at the Cégep de Saint-Laurent. At the same time, he began a program at Concordia University in andragogy. “It’s the science of adult education. To put it simply, she is thinking about how to teach a person who doesn’t have to be with us. »
He’s running out of money, so he starts a second program to get more loans. “I chose politics and I loved it. We read The Republic of Plato, The prince by Machiavelli and works on the history of Quebec and Canada. »
Two years later, the Saint-Laurent CEGEP hired him full-time. He will teach all the physics courses there in addition to creating a new astronomy course without calculations, intended for students in arts and humanities. Because everyone should share the wonder of discovering the cosmos.
In 2011, the federal elections brought him back to politics. He sees the meteoric rise of the New Democratic Party (NDP) in the polls. The dates are still fresh in his memory. “On April 25, it was announced that the NDP could win more seats than the Bloc had ever had. People refused to believe it. However, the numbers were clear. »
He experienced the same thing during the American presidential election in 2016.
For fun, I compiled the national and state polls in an Excel file. The probability of a Trump victory was around 40%. And yet I heard analysts say it was unlikely. I said to myself: either they don’t understand the figures, or they don’t want to believe them.
Philippe J. Fournier
The professor felt he could be useful. In the United States, there were references like Nate Silver. In English Canada, Éric Grenier was active. And of course, academics excel here as elsewhere in this work. But Quebec lacked a popularizer reaching a large audience.
He therefore launched the site Qc125.com. Since then, the phone has never stopped ringing. He now does seat projections for the entire country. “Lately, I have been busy with campaigns in Manitoba and Saskatchewan, among other things. My head is in the Prairies…”
Are there too many surveys? Philippe J. Fournier considers the question strange. Would we like to know less about our society?
PHOTO ROBERT SKINNER, THE PRESS
A coffee with Philippe J. Fournier
First of all, a poll is more serious than a vox pop. Then, all political parties do it. This information should also be known to the public. As long as the polls do not overshadow the coverage of the issues, I see nothing wrong with it, on the contrary.
Philippe J. Fournier
If there is a problem, it comes from errors of interpretation. He gives the example of an old headline which still discourages him. “It was written that after having declined, the PQ made up lost ground. Yet the change was one percentage point. Nothing had changed! »
So the variation was within the margin of error? Not really, replies the physics teacher. “It’s a precise statistical concept. To understand, imagine a population of 100 people. In a bowl, place 100 marbles. Then, we draw at random. Each person has an equal probability of being fooled. In this context, we could calculate a margin of error. The larger the proportion of balls we pick, the more the margin of error decreases. »
In an internet survey, it’s different. Not all individuals have an equal chance of being surveyed. We only have the point of view of Internet users who agree to respond. This is why we write that the sample would have, for example, a margin of error of plus or minus 4%, “if it were probabilistic”.
“But it’s not that serious,” he adds. Let’s imagine the distance between Montreal and Quebec. We don’t need to know it to the nearest centimeter. The current estimate is good enough for our purposes. »
The proof, he says: the quality of the polls during the 2022 electoral campaign. “Overall, they were good. »
For Philippe J. Fournier, these surveys serve as raw material. It integrates them into its model, which relies on previous elections to convert national and regional polls into estimates of voting intentions in each constituency. He refines his model by adding sociodemographic data such as age, income and education. Then it projects a winner for each district.
His computer makes some 50,000 simulations based on the surveys. The operation takes about an hour. “For the federal government, there is more data, so it lasts between two or three hours. »
If the parties and analysts – mea culpa here – studied its figures better, they would avoid errors.
After the last debate in 2018, François Legault flew to Chibougamau. “It was said that he avoided crowds, that he played it safe. In fact, he had polls showing the race in the Ungava constituency was close, so he tried to move the needle. »
Parties sometimes share their internal polls with him so that he can integrate them into his model. He is wary of selection bias. “The parties never talk to me about their bad polls! » He nevertheless checks the methodology. If it is credible, he takes it into account.
As at the observatory, he still has fun connecting the little data points. For this pleasure of seeing take shape under his calculations what was hidden in front of him and which he was unable to see with the naked eye.
Questionnaire without filter
Coffee and me: In a hurry in the morning: espresso capsule, often black. On a quiet morning: strong filter coffee, a touch of soy milk. I like to vary. I don’t have a single recipe (but never cow’s milk).
A book everyone should read: A classic : A brief history of time by Stephen Hawking. A recent one: The End of Everything (Astrophysically Speaking) by Katie Mack. Unlike the writings of politicians or religious and spiritual leaders who want to make us feel big and important, reading about cosmology puts us in our place. Our small and insignificant place.
My latest landmark book: The comic book collection Doonesbury by Garry Trudeau during the presidency of George W. Bush. To laugh and cry about it. One page after another. Great visual art and impactful texts.
A person who inspires me: I admire journalists, real ones. Without them, democracy is impossible. There would only be those who want power and their supporters who want to submit to it.
People living or dead with whom I would like to have dinner: I would like to answer a great scientist like Marie Curie or one of the first astronauts on the lunar missions… but I think it is better not to meet our heroes. They will definitely disappoint us.
A historical event that I would have liked to attend: The great debate in the world of astronomy. Is the Milky Way the only large galaxy in the Universe or are these “spiral nebulae” that we observe in the sky indeed galaxies separate from ours? We learned soon after that the universe is much, much bigger than we thought.
Who is Philippe J. Fournier
- Holder of a bachelor’s degree in physics from McGill University, a master’s degree in astrophysics from Laval University and a certificate in adult education from Concordia University.
- Professor of physics and astronomy at the Cégep de Saint-Laurent, in Montreal.
- Creator of the site Qc125.com and contributor to various media, including the magazine News.