Carolina Torreblanca
@carotorreblanca.bsky.social
140 followers 84 following 9 posts
PhD in Politics (NYU) || Postdoc, PDRI-DevLab @ UPenn || Letterboxd fanatic
Posts Media Videos Starter Packs
Reposted by Carolina Torreblanca
guygrossman.bsky.social
@carotorreblanca.bsky.social , Will Dinneen, and my paper entitled "Political Science Under Pressure: Competition and Collaboration in a Growing Discipline, 2003-2023" has been (conditionally) accepted at @poppublicsphere.bsky.social.

This paper has been a labor of love. osf.io/preprints/os...
carotorreblanca.bsky.social
If you made it this far, good for you. True commitment. In that case, you probably want some more - check out the full paper: osf.io/rqzsg
OSF
osf.io
carotorreblanca.bsky.social
How do novelty and focus affect success? More novel and topically narrow papers get better citations than peers in the same outlet/year, but aren't more likely to land in top 20 journals. They struggle for prestige but become hits!
carotorreblanca.bsky.social
Using STM models on paper text, we measure topical novelty and focus. Co-authored papers are more topically novel than single-author ones, even accounting for author characteristics
carotorreblanca.bsky.social
What research designs are these quant papers using? We can answer that question with our insane data! Our dictionary+GPT model shows design-based approaches growing, though still a minority. Survey experiments have REALLY taken off while IV methods have seen a rise and fall in popularity
carotorreblanca.bsky.social
Out methods have shifted with out size. Our trained model shows PS becoming increasingly quantitative. In 2003, just 45% of papers used quantitative methods; by 2023, that grew to 57% - For Top 20 journals, the shift is even more extreme: from 70% to 85% in 2023
carotorreblanca.bsky.social
Such growth isn't from productivity gains (minimal improvements there). Instead, expansion comes from new authors publishing in political science at increasing rates each year
carotorreblanca.bsky.social
This volume increase comes mostly from co-authored work. Since 2021, most published papers are co-authored. Teams have grown larger too—note the narrowing gap between blue and green line
carotorreblanca.bsky.social
First, we find the discipline has exploded in size. Almost 3x more political science papers published in 2023 than in 2003, and the average journal publishes nearly twice as many articles per year...
carotorreblanca.bsky.social
¡Hola, mundo! My first, ehm, skeet (?) on here is about a new working paper from @guygrossman.bsky.social Will Dinneen and I. It’s on Political Scientists’ favorite topic: Political Science. We look at 140,000 PS papers published in 174 journals over the last two decades. Some of the highlights: