Lisa Argyle
LightNews LightNews
lpargyle.bsky.social
Lisa Argyle
@lpargyle.bsky.social
340 followers 210 following 13 posts
Associate Professor of Political Science at Purdue. Study: Political Conversation, Polarization, Computational Social Science, LLMs
Posts Replies Media Videos
Reposted by Lisa Argyle
natcomputsci.nature.com
📢 @lpargyle.bsky.social, @ethanbusby.bsky.social and colleagues discuss in a Perspective the challenges and opportunities related to scientific inference with LLMs in the political and social sciences. www.nature.com/articles/s43... #cssky #PoliSci

🔓https://rdcu.be/eGwK9
Arti-‘fickle’ intelligence: using LLMs as a tool for inference in the political and social sciences - Nature Computational Science
Large language models are increasingly important in social science research. The authors provide guidance on how best to validate and use these models as rigorous tools to further scientific inference...
www.nature.com
September 15, 2025 at 9:33 PM
lpargyle.bsky.social
Our paper! @ethanbusby.bsky.social Alex Lyman, Bryce Hepner, Josh Gubler, David Wingate
thomasdavidson.bsky.social Thomas Davidson @thomasdavidson.bsky.social · Aug 1
Lyman and colleagues explore the trade-offs between using instruction-tuned models and base versions of LLMs for downstream tasks, emphasizing the need to better understand model training journals.sagepub.com/doi/abs/10.1...
August 1, 2025 at 5:11 PM
lpargyle.bsky.social
Excited to see this excellent collection of papers come out! Many thanks to @thomasdavidson.bsky.social and Daniel Karell
thomasdavidson.bsky.social Thomas Davidson @thomasdavidson.bsky.social · Aug 1
I’m delighted to share that the August 2025 special issue of Sociological Methods & Research on Generative AI is out now. Along 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 ten papers:
A screenshot of the Sociological Methods & Research website showing the special issue title
August 1, 2025 at 5:08 PM
Reposted by Lisa Argyle
ic2s2.bsky.social
Day 1 keynote - Continuing the LLM conversation, Lisa P. Argyle presented “Arti-fickle’ Intelligence: Using LLMs as a Tool for Inference in the Political and Social Sciences.” Which guidelines do you resonate with the most to establish the failure and success of LLMs? #ic2s2
July 22, 2025 at 3:55 PM
Reposted by Lisa Argyle
jmendelsohn2.bsky.social
📣 Super excited to organize the first workshop on ✨NLP for Democracy✨ at COLM @colmweb.org!!

Check out our website: sites.google.com/andrew.cmu.e...

Call for submissions (extended abstracts) due June 19, 11:59pm AoE

#COLM2025 #LLMs #NLP #NLProc #ComputationalSocialScience
NLP 4 Democracy - COLM 2025
sites.google.com
May 21, 2025 at 4:39 PM
Reposted by Lisa Argyle
profsamperry.bsky.social
Are you a PhD student in Political Science or Sociology interested in religion & public life? Apply for a *fully funded* opportunity to join us this summer (June 16-21) as part of the inaugural Wheatley Seminar in Religion & Politics. Apply by March 31. Details here:
wheatley.byu.edu/religionsemi...
Wheatley Seminar on Religion and Politics
wheatley.byu.edu
March 5, 2025 at 8:20 PM
Reposted by Lisa Argyle
jongreen.bsky.social
Ok, thread time.

Closed-ended survey responses are efficient and easy to analyze, but limit what respondents can say. Open-ended responses are useful for letting respondents answer with more depth in their own words (as opposed to yours).
jongreen.bsky.social Jon Green @jongreen.bsky.social · Jan 30
Online/open access in @polanalysis.bsky.social w/ @willrhobbs.bsky.social: a theory and method for inferring attitudes in open-ended survey responses www.cambridge.org/core/journal...
Abstract: Past work on closed-ended survey responses demonstrates that inferring stable political attitudes requires separating signal from noise in “top of the head” answers to researchers’ questions. We outline a corresponding theory of the open-ended response, in which respondents make narrow, stand-in statements to convey more abstract, general attitudes. We then present a method designed to infer those attitudes. Our approach leverages co-variation with words used relatively frequently across respondents to infer what else they could have said without substantively changing what they meant—linking narrow themes to each other through associations with contextually prevalent words. This reflects the intuition that a respondent may use different specific statements at different points in time to convey similar meaning. We validate this approach using panel data in which respondents answer the same open-ended questions (concerning healthcare policy, most important problems, and evaluations of political parties) at multiple points in time, showing that our method’s output consistently exhibits higher within-subject correlations than hand-coding of narrow response categories, topic modeling, and large language model output. Finally, we show how large language models can be used to complement—but not, at present, substitute—our “implied word” method.
January 31, 2025 at 1:56 PM
Reposted by Lisa Argyle
johnholbein1.bsky.social
Gloria in excel sheets deo
December 20, 2024 at 11:53 PM
lpargyle.bsky.social
Love working with the brilliant @shugars.bsky.social!
shugars.bsky.social Sarah Shugars @shugars.bsky.social · Sep 6
Great to see @lpargyle.bsky.social presenting our handbook chapter on innovations in computational political psychology #APSA2024
September 8, 2024 at 3:51 AM
lpargyle.bsky.social
Join us in Puerto Rico to talk AI and Political science! Deadline Friday
lpargyle.bsky.social Lisa Argyle @lpargyle.bsky.social · Jul 22
Computational Political Scientists:
@EthanBusby
and I are organizing a CWC at
@spsanews
on Generative Language Models. Apply by August 16; Conference in Puerto Rico Jan 8 - 11. Please apply and circulate widely! forms.gle/vJoU2ZaMp773... polisky
SPSA 2025 CWC: Generative Language Models in Political Science Research
Apply here for a Conference-Within-a-Conference to be held at the 2025 SPSA Annual Meeting, January 8 - 11 Caribe Hilton, San Juan Puerto Rico. CWC Scope: The topic of the CWC is Generative Language...
forms.gle
August 14, 2024 at 1:32 PM
lpargyle.bsky.social
Computational Political Scientists:
@EthanBusby
and I are organizing a CWC at
@spsanews
on Generative Language Models. Apply by August 16; Conference in Puerto Rico Jan 8 - 11. Please apply and circulate widely! forms.gle/vJoU2ZaMp773... polisky
SPSA 2025 CWC: Generative Language Models in Political Science Research
Apply here for a Conference-Within-a-Conference to be held at the 2025 SPSA Annual Meeting, January 8 - 11 Caribe Hilton, San Juan Puerto Rico. CWC Scope: The topic of the CWC is Generative Language...
forms.gle
July 22, 2024 at 9:08 PM
lpargyle.bsky.social
This is such a a cool service to provide, and a spectacular team to work with!
chrisbail.bsky.social Chris Bail @chrisbail.bsky.social · Apr 24
Science education is so vital. Check out the amazing work of "Science Journal for Kids" which recently covered our research on using AI chatbots to moderate conflict in sensitive political discussions: www.sciencejournalforkids.org/articles/how... Lots of other content created by a small team.
How can AI make online chats kinder? - Science Journal for Kids and Teens
+ Audio version of article + Blackboard version of article + Links to lessons on political discourse and civil debates
www.sciencejournalforkids.org
April 24, 2024 at 5:58 PM
Reposted by Lisa Argyle
bfschaffner.bsky.social
we just released the call for teams for the 2024 Cooperative Election Study! cces.gov.harvard.edu/news/call-te...
Call for teams for the 2024 CES
Announcing the 2024 Cooperative Election Study
cces.gov.harvard.edu
March 4, 2024 at 9:29 PM
lpargyle.bsky.social
This is such a cool and exciting project!
November 14, 2023 at 2:54 PM
Reposted by Lisa Argyle
lukethorburn.com
Recent study by Lisa Argyle, Chris Bail et al. showing that using an LLM to suggest rephrasings in conversations about divisive topics can improve perceived conversation quality and feelings of democratic reciprocity — without any apparent manipulation of political views.

doi.org/10.1073/pnas...
October 12, 2023 at 11:54 AM
Reposted by Lisa Argyle
rossdahlke.bsky.social
🤖 AI can improve the quality of political conversations online without manipulating policy positions 🤖

Great AI-Mediated Communication (AIMC) study from @lpargyle.bsky.social @chrisbail.bsky.social Busby, Gubler, Howe, Rytting, @taylor-sorensen.bsky.social & Wingate

doi.org/10.1073/pnas.2311627120
Treated conversation flow: Respondents write messages unimpeded until one partner receives a rephrasing prompt for the first message longer than four words, and every other conversational turn thereafter. The chat assistant intercepts the treated user’s message, using GPT-3 to propose evidence-based alternative phrasings, while retaining the semantic content. It suggests three randomly ordered alternatives to the author of the message and presents the opportunity to accept or edit any of these rephrasing suggestions or send their original message. Their choice is sent to their partner and the conversation continues. Analysis of semantic content of messages. Panel (A) presents a visualization of the topical distribution of messages sent on the platform. Each point is the semantic embedding of a message; points that are close to each other represent messages that are semantically similar. Messages are clustered with k-means, and clusters are automatically labeled by GPT-4 Conversational quality Panel (A) and democratic reciprocity Panel (B) subgroup means and confidence intervals. Higher values signify greater quality/support.
October 21, 2023 at 5:23 PM
Reposted by Lisa Argyle
chrisbail.bsky.social
Repost if you’ve participated in a Summer Institute in Computational Social Science. Let’s get #SICSS Bluesky going!
October 8, 2023 at 7:49 PM
lpargyle.bsky.social
No better first 🟦 than a new paper! Hot off the press at PNAS: "Leveraging AI for democratic discourse: Chat interventions can improve online political conversations at scale."

Real-time recommendations from an LLM improve divisive conversations.

Open access at: www.pnas.org/doi/10.1073/...
Political discourse is the soul of democracy, but misunderstanding and conflict can fester in divisive conversations. The widespread shift to online discourse exacerbates many of these problems and corrodes the capacity of diverse societies to cooperate in solving social problems. Scholars and civil society groups promote interventions that make conversations less divisive or more productive, but scaling these efforts to online discourse is challenging. We conduct a large-scale experiment that demonstrates how online conversations about divisive topics can be improved with AI tools. Specifically, we employ a large language model to make real-time, evidence-based recommendations intended to improve participants’ perception of feeling understood. These interventions improve reported conversation quality, promote democratic reciprocity, and improve the tone, without systematically changing the content of the conversation or moving people’s policy attitudes
October 3, 2023 at 6:14 PM