Felix Lennert

tack så mycket :)

Thanks :)

Thank you :)

Thank you :)

This also concludes my time in academia (for now).

I am actively looking for industry positions (data science/consulting). If you or someone you know has successfully made this transition and can proffer advice, I'd love to chat.

I tried to add ALT text, some of it got truncated. I did not add ALT text for the French summary as to not confuse anyone who listens to this.

Acknowledgments Note: I’ve listed people in alphabetical order and groups in roughly chronological order. The former serves to avoid inter-personal conflict, the latter to avoid inter-group conflict. As a sociologist, I will happily admit that this thesis is less of a personal achievement but rather the result of being surrounded by excellent people. This section is dedicated to them. Stepping onto the aesthetically challenged campus of the University of Regensburg after a turbulent high school career felt like a breath of fresh air. I cannot thank Melanie Walter-Rogg enough for the life- altering introduction to empirical research methods, the excellent mentorship, and the trust she put in me when letting me off the leash to teach students for the first time. My 5 years there would not have been the same pleasant experience without my community of avid library users. Benedikt, David, Felix, Janik, Ludwig, Martin, Patrick, Sebastian, Stefan, Tarek, Tilman, and Tobias, you changed my perspective on course work from it being a mere chore to rather being a side effect of a nice day with friends. Thank you. Linköping University came with a nicer view and the perfect fit in terms of program. The Master’s program in Computational Social Science provided me with excellent education and I found myself, again, surrounded by great people. The motley crew at the Institute for Analytical Sociology taught me to think out of the box and then back it up with rigorous analyses. A special thanks goes to Marc Keuschnigg for taking me on as his mentee, and eventually Research Assistant, for providing perspective when required, and challenging my thinking whenever needed. This thesis would not have been written if it had not been for your example of how interesting and fulfilling research as a profession can be. Norrköping also provided a temporary home to a great group of people. Adedapo, Gustav, Jan, Pablo,

Paul, and Simona, thank you for making the dreary Swedish winter more bearable and the darker days brighter. My academic journey would have ended there, had it not been for Étienne Ollion. He trusted my abilities and extended them in unforeseeable ways in Paris. Thank you for introducing me to a whole new level of rigor, teaching me invaluable soft skills, and giving me the room and possibilities to thrive. I thoroughly enjoyed our collaboration, be it the Summer Institute in Computational Social Science, the NLP & Society seminar, or the countless hours that we spent working through my scattered ideas and sub-par writing. It is only now in hindsight that I come to slowly appreciate how your mentorship has formed my thinking and the way I navigate the world. My time at CREST would have not been the same without my colleagues there. Thank you, especially, Felix, Luis, and Sofian, for helping me navigate the particularities of life in France and making me feel less alienated in the big city. Annina, Arnout, Emma, Germain, Jeanne, Julia, Mathis, Matteo, Maxime, Patrick, Pedro, Théo, Vincent, and Yasmine, I hold our time in high esteem and am grateful that we essentially never discussed work but rather focused on the fun parts of life. I also need to thank Emiliano Grossman and Samuel Coavoux for serving on my Comité de Suivi. You made these meetings never feel like the chore that they can be but rather a productive check-in that helped steer the projects in a more productive direction. Finally, Frédéric Gonthier for his thorough engagement with this manuscript and for serving as the Président on my committee. I also had the opportunity to visit Duke University twice. This is all thanks to Professor Stephen Vaisey. Thank you for taking me on as a Visiting Scholar, for your great hospitality, and for providing an excellent environment for intellectual growth. The Worldview Lab is a unique place and I will always think of my stays there very fondly. This is especially cou…

Julius, Leonie, Sascha, and Stephan for making these days feel shorter and (sometimes) the nights a bit longer. I also thank my students for being such a great audience for my ramblings on text as data, I hope you could learn something. A heartfelt thanks goes to my co-authors for bearing with me when I switched analysis plans around, and for keeping me in check with writing and analysis schedules. Alexi, Anastasia, Elida, Maans, Miriam, Patrick, and Väinö, thank you very much for genuinely fun and inspiring meetings and great conference visits. Furthermore, there are people outside the ivory tower that academia can be that have made this jour- ney easier. A special thanks goes out to Sven for showing me life outside the academy, for repeated hosting in Los Angeles when I was craving Vitamin D, and weekly calls during what can definitely be considered the loneliest time of my life. In addition, I thank Jan, Johannes, Lars, Selden, and Tim for the excellent bike packing trips against the backdrop of the most magnificent sceneries. Thank you, Steven, for being an En- glish speaker in a French environment and for exploring the beauties of the French back country together. Thank you to the folks at home for making my stays back there like actual vacations and grounding me. I’m unsure if I could have made it through the final stretch without the unwavering support of the world’s kindest person, most-loving cat mom, my best friend, partner, intellectual inspiration, and wife, Alexandra. Thank you for your 5 A.M. wake-up calls and your kind – yet rigorous – feedback on my meandering sentences. I cannot wait to come home to Boston, the cats, and you now that this ordeal is over. Thank you also to my new American family, Lawrence, Nora, and Ritchie, for your support in transitionary times, excellent hosting, and great feedback on my English. Finally, I need to extend my special gratitude to my family. My parents for the unconditional support in all my endeavors, my brothers …

Abstract: This dissertation investigates three facets of political polarization – ideological, affec- tive, and lifestyle – in European multi-party democ- racies. While extant research has overwhelmingly focused on the United States’ two-party system, this serves as a test for the generalizability of polariza- tion theory across party systems. The first chap- ter, combining 25 years of French survey and news- paper data, demonstrates that media reflect opin- ion division on established issues while elite cues shape emerging issue alignments. The second chap- ter, using bi-weekly survey data and automated classification of social media messages during Ger- many’s 2021 elections, shows that polarized elite rhetoric directly predicts mass affective polarization in real-time, with effects moderated by strength of party identification. The third chapter, examining Swedish Twitter users’ following behavior, extends lifestyle polarization theory to multi-party systems, revealing that “cultural bundles” of everyday inter- ests align with broader ideological blocs rather than specific parties. Across these chapters, I demon- strate how digital traces of human behavior com- bined with representative survey data can capture and advance polarization theory in naturalistic set- tings.

It gives me tremendous pleasure to announce that I managed to defend my Ph.D. thesis at ENSAE, IP Paris, last week.

I forgot to take the customary pictures with my committee, so please enjoy the acknowledgments.

You can find a summary of the thesis in French and English as well.

IIRC, in Europe this would require explicit consent by every student who attends the class and therefore might be visible or audible in the recording. At least that was the case during the pandemic days.

Not sure what the rules are for you though of course.

“Max Weber, widely known for his outstanding grills, was also a decent social theorist” (unfortunately I can’t remember where I read this)

Studying a social science will really change how you walk the earth and see society

Me, after 11 years of studies: LOL, Coleman cooler, just like the boat/bathtub

SICSS-Paris is on!
Today, over 40 scholars from all over Europe (and beyond) gathered at @crestumr.bsky.social for 2 weeks of courses and talks about digital methods and text analysis.
And in the evening, @eunjikim.bsky.social gave the first talk on influencers and political change.

Screenshot of the SICSS Paris logo

Next week, CREST is hosting another edition of the Summer School in Computational Social Science (SICSS).

Check out the speakers and participants here: sicss.io/2025/paris/p... and some here:

@bartbonikowski.bsky.social @eunjikim.bsky.social
@austin-van-loon.bsky.social @dirkhovy.bsky.social

One academic and one imposter (left) at an academic conference standing in front of their poster

Thank you for coming out and vibing with us and our poster at PAA. The poster session is over now. So if you haven’t stopped by, please feel guilty.

Download the preprint and all your sins shall be forgiven (doi.org/10.31235/osf...)

Sarah Williams, Associate Professor of Technology and Urban Planning at MIT, will be a keynote speaker for #IC2S2 2025 in Norrköping, Sweden.
🎙Urban policy and big data for public good.
Learn more: www.ic2s2-2025.org/keynotes/

Post a fictional character who you think would be an antivaxxer.

Hard mode: No villains

IC2S2 2025: As many of you have seen, the submission system was temporarily down. It is running again and we extended the deadline to include Feb 26. Please excuse the inconvenience

Among other great locations, Paris will also hold a Summer Institute.

This year's focus will be on LLMs and GenAI.

Join us!

Prof. Arnout Van de Rijt of Sociology at the European University Institute will be a keynote speaker for #IC2S2 2025 in Norrköping, Sweden. 🎙The long-term effects of good and bad luck on life course achievement.
Learn more: ic2s2-2025.org. Stay tuned for more updates!

Interested in a free seminar? To mark our first semester on Bluesky, we’re offering a free seminar from our Spring offerings (through May 31) to one lucky winner. Follow us and repost this message by Monday, February 10 to enter. Winner revealed on Feb. 14—may the odds be in your favor!

IC2S2 2025 has received 23 tutorial proposals—thank you for your excellent proposals! Our review team is at work, and we look forward to announcing the lineup soon. Stay tuned: www.ic2s2-2025.org

🚨 #IC2S2’25 Call for Tutorials deadline is just around the corner—Jan 17, 2025
Don’t miss your chance to shape the computational social science conversation in Norrköping, Sweden.
Submit your tutorial proposal now: www.ic2s2-2025.org/submit-tutorial
Keynote announcements are coming soon!

NEW WORKING PAPER: CREST's @felixlennert.bsky.social and @ppraeg.bsky.social conduct a machine-learning text analysis of intergenerational mobility perceptions in Germany, Sweden, and the United Kingdom

Now on @socarxiv.bsky.social: doi.org/10.31235/osf...

I pressed some buttons on my keyboard, had a couple of meetings with Alexi Gugushvili and @ppraeg.bsky.social, and this came out: osf.io/preprints/so...

Will probably write a more extensive description of this once I have wrapped up the teaching of this semester and reorganized my life.

Doors "open" `if lubridate::today() <= date_in_question` and, hence, remain clickable. The full tibble will be shared with her once Christmas has arrived.
Code available upon request (sans tibble).

The base of it is basically a tibble containing the different days and links. These could literally lead to anything. I chose to populate it with links to cute and fun activities for when distance is not as big as the Atlantic.

Are you a nerd (like me) and want to shorten your loved one's/ones' wait until Christmas? Two unacknowledged AI helpers and I (the triumvirate of code taming) teamed up to program a little Shiny Advent calendar.

Reine E-Autos gelten ab 2 Tonnen als SUV, Quelle: www.sueddeutsche.de/panorama/par...

Dear all, coming Wednesday, 17/1, at 17.15, we will welcome Alex Kindel (Sciences Po médialab) to our NLP & social sciences seminar. He will talk about his recent work on word embedding association tests. You can follow online. Find more information here: www.css.cnrs.fr/nlp-social-s...

Interesting. No worries! Glad it’s worked :-)

And this is the full URL: bookdown.org/f_lennert/to... – it works fine for me.
The raw qmd files can be found in this Github repo: github.com/fellennert/t...