Sean Westwood
@seanjwestwood
Followers
2K
Following
1K
Media
67
Statuses
829
Political Scientist at Dartmouth. Director Polarization Research Lab. According to Nate Silver: "Boring. Can't model for shit."
Hanover, NH
Joined October 2009
🚨New working paper🚨 We find that issue-specific bipartisan appeals credibly signal policymaking intentions We show why bipartisan lawmaking persists amid congressional conflict: progress comes through issue-specific pathways rather than broad consensus between Ds and Rs 🧵👇
2
6
21
Most normie taxpayers care about teaching and hours in the classroom, and don’t see value in the majority of academic research Professors are paid to teach and if they teach so little, then they need to be paid less or teach more Brag about your free time at your own risk!
GOP reps and state legislators filing away all of the humblebrag QTs (“I get paid so much to do so little”) in their “Reasons to Defund Higher Ed” folders
3
2
41
"Unlike scientific inquiry, feminist studies rarely initiate self-correction when research and data demand it. The field tends to be isolated within an ideological echo chamber unfit for academia. Ideology reigns over empirical inquiry."
73
359
2K
Excited to share our new paper published in PNAS (joint with @SemraSevi and Don Green)! AI can enhance political knowledge and provide balanced information about politics with proper guardrails and vetted sources (e.g., party platforms). https://t.co/xKA0cFyATW
2
20
47
If true this would be a big deal, but I suspect if they had a 100% solution the details would be public.
0
0
6
I hear @CloudResearch is promoting that they have solved the AI respondent problem and are using my work to sell the solution. I have no relationship with them and they have not provided *ANY* evidence of efficacy. Be very skeptical. Again, I encourage them to post data or a
2
14
52
Politicians who later become congressional leaders trade stocks like everyone else ... until they ascend to power. After ascension, their portfolios beat peers by 47 (!!!) percentage points a year through trades timed around bills and firms that later get government contracts.
9
248
937
Good thread. I think the opt-in survey panel had a very good 20 year run, but the paradigm may need to change. Much of my recent work has built custom sampling frames:
cambridge.org
The Politics of Small Business Owners - Volume 55
0
10
50
How far has the credibility revolution actually reshaped the discipline? In a new working paper w/ @caro_whitetower @william_dinneen & Guy Grossman, we use GPT-4o to code 91,632 articles from 174 political science journals (2003–2023) and track research designs, transparency
8
109
315
The most damning evidence of this is the proliferation of online classes in the era of LLMs
1
1
5
The reddits run by survey panelists are terrifying. These are people bragging about completing 10,000+ tasks and making a living on Prolific.
4
4
80
Watch my tool simulate a human entering text on a survey. Real headwinds against survey work.
Large language models can answer surveys and pass the tests to check that a respondent is human
3
11
58
Some four-year college degrees are not worth the cost. STEM or gov/econ at your state school is still affordable and extremely worth it for most.
Poll: In a dramatic shift, Americans no longer see four-year college degrees as worth the cost https://t.co/DuHGLzBTWp
13
7
93
Writing your students grad school rec letters? Encourage them to also apply for our political science predoc position at UC Berkeley! https://t.co/BLBOCOlcg4
aprecruit.berkeley.edu
University of California, Berkeley is hiring. Apply now!
0
3
8
When your kid says they are a bad test taker your first question should be: how much did you actually study for that test?
One of the most common myths in education is the myth of the “bad test-taker.” In a study of college students, the majority (56%) identified themselves as bad test-takers. It’s a self-defeating myth that discourages effective studying and the use of test results as a yardstick.
5
16
93
I don't think it is possible to say that this is either predictable or inevitable. It is tragic.
Tragic and despicable but totally predictable and inevitable….we are relearning the lessons of Baghdad
0
1
5
NEW PAPER w/ @CSElmendorf & @j_kalla: An under-appreciated reason why voters oppose dense new housing, especially in less-dense neighborhoods: they think it looks ugly and want to prevent that, even in other neighborhoods. Some of what we think is NIMBYism might not be! 🧵
33
118
589
Strong math skills matter a lot. Weak math skills close doors.
The college majors with the best return on investment are also the ones you can’t get into without strong math skills?
6
26
149
The academic system is cratering. Grades are inceasingly disconnected with performance or knowledge.
One of the Harvard report’s most compelling findings is that students almost universally speak about grades in terms of how much effort they put in. If they spend a lot of time studying and do all of the work asked of them, they believe, they should get an A.
25
123
686
As @seanjwestwood's terrifying new PNAS article demonstrates, LLMs can now pass almost every attention check, mirror personas, stay consistent across pages, and systematically bias responses in the aggregate. So here’s a different angle: verify physical presence, not text.
3
13
34