Shamira Gelbman

Tonight's fun

How it started, how it's going 🤷‍♀️

need crowdsourcing help #polisky - if your department is not admitting PhD students this cycle, please respond with the department

i already know my department, GW, is not admitting students

Letters like this hit different in the ChatGPT era 🙃
#archives #LadyLobbyists

Presentation title slide: "Taking the 'Lipstick Lobby Brigade' Seriously: Toward a Study of World War II-Era Women Lobbyists" 1940s lipstick ad text: "In Defense of Glamour: Patriot Red. The Brave New Shade" Shamira Gelbman | Wabash College | gelbmans@wabash.edu

I'm presenting today at #ESS2026 👀

I'll introduce my new project on WWII-era women lobbyists drawing on initial findings and questions raised through archival work with the @ncjw.bsky.social Washington Office Records

3:30 PM in Salon 2 in the Marriott at Metro Center

Road trip pause for ~5 weeks of conferencing and other nerdy fun in DC 🤓

Greensboro, NC #polisciroadtrip 🚗

Oh autocorrect, I wasn't trying to type "Fulkerson" 🙄🤦‍♀️

User prompt: I would like a summary of why AI is not better at social science research than most professors with PhDs Claude response: AI systems face several structural limitations that constrain their usefulness in social science research relative to experienced researchers: Methodological judgment Social science research requires nuanced decisions about research design, variable operationalization, and appropriate analytical frameworks. These judgments depend on deep familiarity with a discipline's epistemological debates and contested assumptions — knowledge that PhD researchers develop through years of supervised practice and peer engagement. Contextual and cultural knowledge Interpreting social phenomena requires understanding historical context, institutional dynamics, and cultural meaning that resist reduction to patterns in text data. Fieldwork, interviews, and ethnographic observation generate tacit knowledge that AI systems cannot replicate. Causal inference Identifying causal mechanisms in social systems — distinguishing confounders, selecting instruments, arguing for identification strategies — requires theoretical reasoning about social causation that goes beyond statistical pattern recognition. Original data collection Much social science advances through novel primary data: surveys, experiments, archival research, interviews. AI systems work with existing data and cannot independently design and execute data collection strategies responsive to a specific research question. Theoretical contribution Advancing theory requires identifying gaps in existing literature, constructing novel conceptual frameworks, and situating findings within disciplinary conversations. This demands genuine understanding of what is already known and why it matters. Critical evaluation Assessing the validity of competing claims, identifying flaws in published work, and exercising skepticism about methodological choices requires judgment that AI systems apply inconsistently.

Claude had a really interesting to answer when I asked the negative question

Beckley, WV #polisciroadtrip 🚗

Charleston, WV #polisciroadtrip🚗

WV, KY, and OH 👀 #polisciroadtrip 🚗

NEW SPEAKERS ANNOUNCED!
Join us for the Peopling Politics Symposium on March 5 in DC, featuring Rep. Greg Casar, former USTR Katherine Tai, and former Asst. AG Kristen Clarke, plus brilliant panelists across academia, government, and civil society!

RSVP Here: www.eventbrite.com/e/bringing-t...

Does he not remember the donut shop?

Teddy would like a word

These are “low enrollment” majors but the classes are regularly at capacity. They changed the funding structure a few years ago here, switching the allocation from course enrollments to majors then blamed these departments for a structural, top-down change. This is entirely a political choice.

At least they're trying 🤷‍♀️

Finished the Duolingo Russian course 🎉🎉🎉

Maybe German next? IDK 🤷‍♀️

👀

... So finally I had to break the conversation off by saying, 'Professor, when it comes to the theory of relativity, you are the professor, but when it comes to the garment industry I am the professor.'"

Dubinsky's account may be more on the nose: "For all his brilliance, Einstein was a simple man in matters of this kind ...

When I see mention of Einstein, I think of this story (as told by John W. Edelman in his autobiography, Labor Lobbyist), which possibly has some wisdom for bot-Einstein as well 🤷‍♀️

We do talk about this timeline in my #PSC214 Politics of the Civil Rights Movement course -- this video could be useful next time I teach it 👀

Logo featuring the letters "PS" and the phrase "Teacher Symposium" alongside a thermometer with a globe at its center.

Teaching Qualitative Methods in Undergraduate Education.

A symposium from @pspolisci.bsky.social -

https://cup.org/4tDnyNn

i've decided not to listen to the audio of a teacher reporting a kindergartender because frankly I don't want to become the person that that event would turn me into. but one thing I hope we are collectively coming to grips with is what it meant that we have lived among such people this whole time

I might find some use for this in #PSC111 👀

Together with @shamiragelbman.bsky.social , I edited a special section on teaching #qualitative #methods in #polisky to undergrads @pspolisci.bsky.social, currently all ungated on first view:
www.cambridge.org/core/journal...

I'm very excited about the range of contributions - short 🧵1/7

Same boat for me -- there are plenty of other collections to keep my project rolling for the foreseeable future, but it will be a big gaping hole if I can't get there eventually

The reply I got wasn't that informative, but I figured it was something like that. Very disappointing, and I'd definitely be up to help if you/others are taking any action.

Sorry to dredge up this post - I found it after running into the same issue when I emailed them today about a possible research visit 😞