Christian Oswald retweetet
1/🧵 A major update to our paper: "Scaling Reproducibility" w/ @YangYang_Leo.
We move beyond reanalyzing a single design to (almost) full-paper replication!
Paper: bit.ly/repro-ai
English
Christian Oswald
2.5K posts
Christian Oswald
@teamoswald
Research Associate @cissmunich | @PRIOresearch Research School Member | pol. violence & instability | analog player in a digital world | aggressively mediocre
Germany Beigetreten Nisan 2015
1.1K Folgt408 Follower
Christian Oswald retweetet
4w ago I was a Claude Code skeptic. I'm not a coder. None of the use cases were relevant. I managed teams & projects, drowning in email & overdue reminders. So I tried creating tools that would help me and... holy crap.
Now I'm sharing the tools I built:
claudeblattman.com
English
Christian Oswald retweetet
1/ 🐎 Our gift for the Year of the Horse:
An AI-assisted workflow that scales reproducibility in empirical research. w/ @YangYang_Leo
Paper: bit.ly/repro-ai
English
Christian Oswald retweetet
I have a new paper. We look at ~all stats articles in political science post-2010 & show that 94% have abstracts that claim to reject a null. Only 2% present only null results. This is hard to explain unless the research process has a filter that only lets rejections through.
English
Christian Oswald retweetet
Hi all, I've uploaded the 2025 update to my PhD Applied Econometrics slides:
➡️ More on regression & causality
➡️ Dynamic panel data models
➡️ Streamlined diff-in-diff extensions
➡️ More on spillover effects
➡️ Results from new papers on many topics
Link in the original tweet
Kirill Borusyak@borusyak
Hi #Econtwitter, I'd like to share the slides from the PhD Applied Econometrics course I just had the privilege to teach at @AreBerkeley Regression & causality, selection on observables, panel data, IV, RDD --- usual topics but hopefully in a modern way github.com/borusyak/are213
English
Christian Oswald retweetet
This spring I'm GSI-ing graduate formal models. I struggled with proofs at first, and it's not part of math camp so I drafted a guide. I'm sharing it in the spirit of public goods / in case some are interested; I'd be grateful for comments/feedback! dropbox.com/scl/fi/qby8xnf…
English
Christian Oswald retweetet
Christian Oswald retweetet
A Critical Look at the Issue of Causal Inference
@DeatonAngus and Nancy Cartwright raise doubts about the view that randomized controlled trials (RCTs) should be assigned a special status. See sciencedirect.com/science/articl…
For a response by @guido_imbens, see: sciencedirect.com/science/articl…
English
Christian Oswald retweetet
Glad this paper with @guido_imbens is out in the JEP.
The LaLonde paper has had a big impact on my academic journey and continues to teach us about the challenges and possibilities of conducting credible inference using nonexperimental data. aeaweb.org/articles?id=10…
Many thanks to Tim for the tireless editing and suggestions. @TimothyTTaylor
English
Christian Oswald retweetet
New software: "Projoint: The One Stop Conjoint Stop" (with @aaronrkaufman & @YusakuHoriuchi), makes conjoint surveys easier & less biased, including everything from a drag-and-drop web survey design tool to specialized analysis software. Comments welcome. yhoriuchi.github.io/projoint/
English
Christian Oswald retweetet
If you have been curious about which topics are increasing in popularity in econ, @paulgp has a new tool for you!
paulgp.com/econlit-pipeli…
This lets you search over AER and AEJ's + 30k NBER WP!
Very neat!
English
Christian Oswald retweetet
How do promotion prospects drive patterns of repression in China? @baggottcarter, Jonghyuk Lee, and I find that the most promising officials hesitate to use repression for fear of backlash riots, except for in separatist regions journals.uchicago.edu/doi/epdf/10.10…
English
Christian Oswald retweetet
Latest from my colleagues at Forecasting Research Institute (FRI):
Superforecasters still outperform top LLMs.
But for how much longer?
Linear extrapolation from recent ForecastBench data suggests LLMs will surpass top humans in November 2026.
And linear extrapolation is itself a tough benchmark for both humans & LLMs to beat
Forecasting Research Institute@Research_FRI
🔮 When will AI forecasters match top human forecasters at predicting the future? In a recent @cowenconvos podcast episode, @NateSilver538 said 10–15 years while @tylercowen predicted 1–2 years. Who was right? Our updated AI forecasting benchmark, ForecastBench, suggests that Tyler Cowen is more likely to be right.
English
Christian Oswald retweetet
Christian Oswald retweetet
Whoa—my book is up for pre-order!
Model to meaning: How to interpret Statistical & ML Models in #RStats and #PyData
The book presents an ultra-simple and powerful workflow to make sense of ± any model you fit
Note: The web version stays free forever.
tinyurl.com/4fk56fc8
English
Christian Oswald retweetet
"Political violence in democracies: An Introduction" that I coauthored with Ursula Daxecker & Neeraj Prasad is now out & open access, @JPR_journal . We introduce 14 excellent articles and discuss the state of research on political violence in democracies. journals.sagepub.com/doi/full/10.11…
English
Christian Oswald retweetet
I’m delighted to share that the August 2025 special issue of Sociological Methods & Research on Generative AI is out. With my co-editor, Daniel Karell, we put together this issue to build on the conference we organized last year.
Here's a thread on each of the 10 papers:
English
Christian Oswald retweetet
New article out at the American Journal of Political Science:
"The Troubles and Beyond: The impact of a museum exhibit on a post‐conflict society."
With Dr. Elsa Voytas.
Open access here: onlinelibrary.wiley.com/doi/10.1111/aj…
English
Christian Oswald retweetet
ConflLlama: Domain-specific adaptation of large language models for conflict event classification - Shreyas Meher, Patrick T. Brandt, 2025 journals.sagepub.com/doi/full/10.11…
English
Christian Oswald retweetet
We have some updates on our DDD paper.
In our latest version, we have introduced three new applications and an open-source R package to facilitate the usage of all our DDD tools!
🚀R package: marcelortiz.com/triplediff/
🔎Updated paper: arxiv.org/abs/2505.09942
Pedro H. C. Sant'Anna@pedrohcgs
🚨New paper alert🚨 My paper with @marcelortizv on "Better Understanding Triple Differences Estimators" is finally out. 📰 psantanna.com/files/DDD_OVS.… Main msg: Nuances in DDD designs challenge the practice of generally viewing DDD as the diff btw two DiDs. We can do better! 1/
English