Methods, American Politics | Researcher for the Institute of Quantitative Social Science at Harvard
Cambridge, MA
Joined May 2013
- Tweets133
- Following1.1K
- Followers477
- Likes3.4K
Danny Ebanks retweeted
2/2 If you want to understand the dynamic of the midterm election across the nation, read this recent working paper by @DannyCEbanks, @kinggary, and @Jonathan_N_Katz:
gking.harvard.edu/10k/
Also use their amazing election prediction application:
gking.harvard.edu/electit/
Danny Ebanks retweeted
1/2 As we enter the final weekend of the 2026 midterm elections in the US, two things for your reading list. To understanding what is going on in the minds of those who have still haven't voted, take a look at this recent paper with @DannyCEbanks:
frontiersin.org/journals/pol…
A few notes, we give you the chance to build your own forecast going back to 1946. If you think there is another polling error like in 2016, or want to see what the model would say in 1956, try all these possibilities, make your own modeling decisions, and see the numbers change!
New 2026 US House & Senate election forecasts, using all political science knowledge we could systematize, & our statistical methods from “If a Statistical Model Predicts That Common Events Should Occur Only Once in 10,000 Elections, Maybe It’s the Wrong Model” with @DannyCEbanks and @Jonathan_N_Katz. GaryKing.org/electit
We make these forecasts available (including all the input data, methods, and procedures), not only for their possible public interest, but because academics learn by making themselves vulnerable to being proven wrong. We will make no changes to our post-primary forecasts (which do not use polls) but will update two other sets on this site incorporating poll results as new ones appear. Comments on the paper or forecasts are welcome!
We also estimates of candidate effects and the polls to see how candidates help or hurt their party's chances in a particular seat!
Danny Ebanks retweeted
New 2026 US House & Senate election forecasts, using all political science knowledge we could systematize, & our statistical methods from “If a Statistical Model Predicts That Common Events Should Occur Only Once in 10,000 Elections, Maybe It’s the Wrong Model” with @DannyCEbanks and @Jonathan_N_Katz. GaryKing.org/electit
We make these forecasts available (including all the input data, methods, and procedures), not only for their possible public interest, but because academics learn by making themselves vulnerable to being proven wrong. We will make no changes to our post-primary forecasts (which do not use polls) but will update two other sets on this site incorporating poll results as new ones appear. Comments on the paper or forecasts are welcome!
Danny Ebanks retweeted
Dataverse 6.12 is now available, with new API capabilities, performance and security improvements, and a range of bug fixes.
github.com/IQSS/dataverse/re…
#Dataverse #ResearchData #OpenData #OpenScience
Glad to see this out in FirstView!
Currently in FirstView: In “Analyzing Political Text at Scale with Online Tensor LDA,”@SaraKangaslahti, @DannyCEbanks, @JeanKossaifi, @anqi_liu33, @rmichaelalvarez, and @AnimaAnandkumar introduce a topic modeling method that scales linearly to billions of documents.
Glad to see Analyzing Political Text at Scale with Online Tensor LDA finally in print at @polanalysis !
NEW: Analyzing Political Text at Scale with Online Tensor LDA @polanalysis @CambridgeCore|
@SaraKangaslahti-@HarvardUniv @DannyCEbanks-@IQSS @JeanKossaifi-@nvidia @anqi_liu33-@HopkinsEngineer @rmichaelalvarez-@CaltechLCSSP @AnimaAnandkumar-@Caltech share.google/0jQGE8J85KzGUxL…
Forthcoming paper in Political Analysis
The method features mathematically provable guarantees of accuracy and the ability to train the model end-to-end on a local GPU.
Our next project is to take these outputs to improve LLM training and classification performance. More soon!
NEW RESEARCH: @SaraKangaslahti, @HarvardUniv | @DannyCEbanks, @IQSS, Harvard | @JeanKossaifi, @nvidia | @anqi_liu33, @HopkinsEngineer, @JohnsHopkins | @rmichaelalvarez, @Caltech, @CaltechLCSSP | @AnimaAnandkumar, Caltech
Forthcoming in Political Analysis: arxiv.org/abs/2511.07809v1
The model’s theoretical guarantees dovetail with its practical performance so that method can reliably process text on the scale of billions of documents in feasible time, allowing Social Science researchers to vastly scale up their study of text.
Excited to share new published work driven by brilliant (now graduate) researchers Sreemanti Dey and Sarah Hashash and @rmichaelalvarez . We compiled a dataset of social media conversations about election administration to study interactions between election staff and the public.
Our contributions: (a) a flexible, seed-based data-collection architecture for social media; (b) combined JST and temporal modeling to track animosity in near real-time; (c) empirical evidence linking the loser’s effect to spikes in hostile interactions.