Paride Carrara's Avatar

Paride Carrara

@paridecarrara

PostDoc at WICE ERC, University of Latvia | Adjunct Prof. UniMi | PhD UniBo | Interested in party politics, communication, strategic ambiguity & #TextAsData | Travel & photography

287
Followers
1,316
Following
27
Posts
27.10.2023
Joined
Posts Following

Latest posts by Paride Carrara @paridecarrara

Our new paper is out in Party Politics ๐Ÿšจ

10.03.2026 15:18 ๐Ÿ‘ 1 ๐Ÿ” 0 ๐Ÿ’ฌ 0 ๐Ÿ“Œ 0
Post image

๐Ÿšจ New WP w/ @leonardocarella.bsky.social on OSF

doi.org/10.31235/osf...

We usually think that social identities precede preferences

We show the reverse is also true: people update their social identities to match their immigration preferences

Focus: class identity in ๐Ÿ‡ฌ๐Ÿ‡ง + Christian identity in ๐Ÿ‡ฎ๐Ÿ‡น

23.02.2026 10:57 ๐Ÿ‘ 128 ๐Ÿ” 56 ๐Ÿ’ฌ 3 ๐Ÿ“Œ 6

Currently in FirstView: In โ€œPolitical DEBATE: Efficient Zero-Shot and Few-Shot Classifiers for Political Text,โ€ Michael Burnham, Kayla Kahn, Ryan Yang Wang, and Rachel Peng introduce DEBATE, a new open source foundation model for classifying political documents.

06.01.2026 17:35 ๐Ÿ‘ 6 ๐Ÿ” 4 ๐Ÿ’ฌ 1 ๐Ÿ“Œ 1
Post image

๐Ÿšจ 3-Year Postdoc in Political Science at Aarhus University ๐Ÿšจ

Iโ€™m seeking to recruit a postdoc for my @erc.europa.eu research project ๐‘ฌ๐‘ธ๐‘ผ๐‘ฐ๐‘ณ๐‘ฐ๐‘ฉ๐‘น๐‘ฐ๐‘ผ๐‘ด on state-citizen interactions.

Link and more information in second post.

Position Start: Fall 2026
Application Deadline: โ€ผ๏ธ February 5, 2026, 23:59 CET โ€ผ๏ธ

09.12.2025 12:00 ๐Ÿ‘ 75 ๐Ÿ” 69 ๐Ÿ’ฌ 1 ๐Ÿ“Œ 0
Post image Post image

โฐ Iโ€™m happy to share that a major paper from my dissertation was published today (๐Ÿ”— doi.org/10.1017/pan....) in Political Analysis. In the paper, Clemens Lechner and I conduct an extensive validation study on how to measure politiciansโ€™ public personality traits using computational text analysis!

02.12.2025 11:08 ๐Ÿ‘ 26 ๐Ÿ” 8 ๐Ÿ’ฌ 1 ๐Ÿ“Œ 0
BJPolS abstract from a research paper discussing a machine learning approach to analyze political discourses on climate change among OECD, European, and South American countries.

BJPolS abstract from a research paper discussing a machine learning approach to analyze political discourses on climate change among OECD, European, and South American countries.

NEW -

Policlim: A Dataset of Climate Change Discourse in the Political Manifestos of Forty-Five Countries from 1990 to 2022 - https://cup.org/43zzLHc

- @marysanford.bsky.social, Silvia Pianta, @nicolasschmid.bsky.social & Giorgio Musto

#OpenAccess

23.10.2025 09:10 ๐Ÿ‘ 14 ๐Ÿ” 7 ๐Ÿ’ฌ 0 ๐Ÿ“Œ 0
Post image

Excited to share that as of October 6, Iโ€™ve joined the ERC-funded project โ€œWelcoming Immigrants in Central and Eastern Europe: Lessons from Russiaโ€™s Invasion of Ukraineโ€ (WICE) as a Postdoctoral Researcher at the University of Latvia! #postdoc #polisci

19.10.2025 07:58 ๐Ÿ‘ 8 ๐Ÿ” 0 ๐Ÿ’ฌ 1 ๐Ÿ“Œ 0
Preview
Topical Negative Campaigning Under Spatial Pressure: Party-Level Strategies for Attacks Across Multiple Issues This paper investigates how spatial party competition shapes topical negative campaigning during Germany's 2021 federal election. We define topical negative campaigning as attacks on opponents that...

๐Ÿ“ข New Publication Alert!

Thrilled to share our latest article with @jasminriedl.bsky.social & Johannes Steup
๐Ÿ‘‰ Topical Negative Campaigning Under Spatial Pressure: Party-Level Strategies for Attacks Across Multiple Issues
๐Ÿ”— doi.org/10.1080/0964...

06.10.2025 08:26 ๐Ÿ‘ 13 ๐Ÿ” 7 ๐Ÿ’ฌ 1 ๐Ÿ“Œ 1
Post image

You can fool some of the people some of the time, but as campaigns unfold, you can fool fewer of the people.

New paper w/ Derek Beach and Jannik Fenger on framing effects out in @ejprjournal.bsky.social

Open Access: doi.org/10.1017/S147...

02.10.2025 17:31 ๐Ÿ‘ 33 ๐Ÿ” 8 ๐Ÿ’ฌ 1 ๐Ÿ“Œ 0
OSF

๐Ÿ“ฃ MORAL APPEALS IN POLITICAL COMMUNICATION ๐Ÿ“ฃ
New version of @twidmann.bsky.social and my working paper answering:
* Have moral appeals increased over time?
* Is the tendency to moralize ideologically patterned?
* Are some topics consistently more moralized than others?
osf.io/preprints/os...

15.09.2025 07:46 ๐Ÿ‘ 46 ๐Ÿ” 13 ๐Ÿ’ฌ 2 ๐Ÿ“Œ 1
We present our new preprint titled "Large Language Model Hacking: Quantifying the Hidden Risks of Using LLMs for Text Annotation".
We quantify LLM hacking risk through systematic replication of 37 diverse computational social science annotation tasks.
For these tasks, we use a combined set of 2,361 realistic hypotheses that researchers might test using these annotations.
Then, we collect 13 million LLM annotations across plausible LLM configurations.
These annotations feed into 1.4 million regressions testing the hypotheses. 
For a hypothesis with no true effect (ground truth $p > 0.05$), different LLM configurations yield conflicting conclusions.
Checkmarks indicate correct statistical conclusions matching ground truth; crosses indicate LLM hacking -- incorrect conclusions due to annotation errors.
Across all experiments, LLM hacking occurs in 31-50\% of cases even with highly capable models.
Since minor configuration changes can flip scientific conclusions, from correct to incorrect, LLM hacking can be exploited to present anything as statistically significant.

We present our new preprint titled "Large Language Model Hacking: Quantifying the Hidden Risks of Using LLMs for Text Annotation". We quantify LLM hacking risk through systematic replication of 37 diverse computational social science annotation tasks. For these tasks, we use a combined set of 2,361 realistic hypotheses that researchers might test using these annotations. Then, we collect 13 million LLM annotations across plausible LLM configurations. These annotations feed into 1.4 million regressions testing the hypotheses. For a hypothesis with no true effect (ground truth $p > 0.05$), different LLM configurations yield conflicting conclusions. Checkmarks indicate correct statistical conclusions matching ground truth; crosses indicate LLM hacking -- incorrect conclusions due to annotation errors. Across all experiments, LLM hacking occurs in 31-50\% of cases even with highly capable models. Since minor configuration changes can flip scientific conclusions, from correct to incorrect, LLM hacking can be exploited to present anything as statistically significant.

๐Ÿšจ New paper alert ๐Ÿšจ Using LLMs as data annotators, you can produce any scientific result you want. We call this **LLM Hacking**.

Paper: arxiv.org/pdf/2509.08825

12.09.2025 10:33 ๐Ÿ‘ 303 ๐Ÿ” 106 ๐Ÿ’ฌ 6 ๐Ÿ“Œ 23

This is a great example of how the use of innovative data can bring us a step further with old research questions

12.09.2025 09:40 ๐Ÿ‘ 12 ๐Ÿ” 6 ๐Ÿ’ฌ 1 ๐Ÿ“Œ 0
Preview
The politics of seeking and avoiding discourse in parliament ELIAS KOCH, ANDREAS KรœPFER

๐Ÿ“˜ 64.4

Using a ๐Ÿ‡ฉ๐Ÿ‡ช dataset of attempted and successful exchanges in the German Bundestag ๐Ÿ›๏ธ (1990โ€“2020), @eliaskoch.bsky.social & @ankuepfer.bsky.social finds MPs with opposing views seek debate more often, but there is rarely action๐Ÿ˜ฑ

#Politics #Parliament

Read more:

12.09.2025 07:01 ๐Ÿ‘ 3 ๐Ÿ” 2 ๐Ÿ’ฌ 0 ๐Ÿ“Œ 0

Check out this new blog post @danielcaseycbr.bsky.social and I wrote about using e-newsletters from Australian and UK Members of Parliament as data. #polisky

11.09.2025 09:11 ๐Ÿ‘ 4 ๐Ÿ” 3 ๐Ÿ’ฌ 0 ๐Ÿ“Œ 1
Post image

Excited that my first PhD paper is published in Journal of Elections, Public Opinion and Parties!๐ŸŽ‰

๐Ÿงต @wurthmann.bsky.social & I examine: Do They Know What They Represent? Parliamentary Candidatesโ€™ Perceptions of Their Own Partyโ€™s Positions

๐Ÿ”— doi.org/10.1080/1745...

11.09.2025 11:35 ๐Ÿ‘ 64 ๐Ÿ” 20 ๐Ÿ’ฌ 3 ๐Ÿ“Œ 1
Post image

This Friday (12.09), we kick off the academic year at the #HotPoliticsLab with our very own (and recently appointed!) professor Gijs Schumacher!

He will present his work on emotions and political behavior. Join us in the Common Room (REC-B9.22) or online via teams.microsoft.com/dl/launcher/...

10.09.2025 13:28 ๐Ÿ‘ 4 ๐Ÿ” 3 ๐Ÿ’ฌ 0 ๐Ÿ“Œ 0
Post image Post image

The @crestumr.bsky.social AI & Social Sciences seminar resumes.

Wednesday 17th at 5pm (CET), we will receive Pr. @emilymbender.bsky.social . She will talk about her latest book, *The AI Con*, co-authored with Dr. @alexhanna.bsky.social.

Join us for what will certainly be a lively conversation.

09.09.2025 11:13 ๐Ÿ‘ 31 ๐Ÿ” 17 ๐Ÿ’ฌ 2 ๐Ÿ“Œ 0
Post image Post image

Had an amazing time at the #SISP in Naples! Presenting my work and exchanging ideas was truly stimulating.

06.09.2025 18:01 ๐Ÿ‘ 3 ๐Ÿ” 0 ๐Ÿ’ฌ 0 ๐Ÿ“Œ 0
Preview
Sage Journals: Discover world-class research Subscription and open access journals from Sage, the world's leading independent academic publisher.

New POPPA article out in Party Politics by Andrej Zaslove, @robert-a-huber.bsky.social, @mauritsmeijers.bsky.social:

โ€œThe State of Populism: Introducing the 2023 Wave of the Populism and Political Parties Expert Survey."

journals.sagepub.com/doi/10.1177/... 1/2

04.09.2025 10:22 ๐Ÿ‘ 5 ๐Ÿ” 3 ๐Ÿ’ฌ 1 ๐Ÿ“Œ 0
Post image

๐Ÿ“ข Talk Announcement

"ActiveTigger: A Collaborative Text Annotation Research Tool for Computational Social Sciences", by @emilienschultz.bsky.social , Etienne Ollion and Paul Girard.

๐Ÿ“œ pretalx.com/pydata-paris-2025/talk/ZHNCF7
๐Ÿ“… pydata.org/paris2025/schedule
๐ŸŽŸ pydata.org/paris2025/tickets

22.08.2025 07:00 ๐Ÿ‘ 4 ๐Ÿ” 2 ๐Ÿ’ฌ 0 ๐Ÿ“Œ 1
Moral Foundation Measurements Fail to Converge on Multilingual Party Manifestos | Political Analysis | Cambridge Core Moral Foundation Measurements Fail to Converge on Multilingual Party Manifestos

New publication, out in Political Analysis:

There is an increasing array of tools to measure facets of morality in political language. But while they ostensibly measure the same concept, do they actually?

I and @fhopp.bsky.social set out to see what happens.

19.08.2025 07:52 ๐Ÿ‘ 34 ๐Ÿ” 14 ๐Ÿ’ฌ 3 ๐Ÿ“Œ 1
Post image

Why do parties do what they do? Excited to share my 2nd dissertation paper, just published in @jeppjournal.bsky.social
In the paper, I argue that parties seek internal unity and try to keep the team together: www.tandfonline.com/doi/full/10....

1/11 ๐Ÿงต

01.08.2025 07:52 ๐Ÿ‘ 86 ๐Ÿ” 27 ๐Ÿ’ฌ 6 ๐Ÿ“Œ 5
Post image

Social scientists have skills and knowledge that are in demand across a range of sectors leading to a variety of interesting and rewarding careers. Explore our interviews with social scientists to see where a career in social science could take youโžก๏ธ acss.org.uk/category/im-...

18.07.2025 11:22 ๐Ÿ‘ 6 ๐Ÿ” 7 ๐Ÿ’ฌ 0 ๐Ÿ“Œ 0
Abstract

Do policy priorities that candidates emphasize during election campaigns predict their subsequent legislative activities? We study this question by assembling novel data on legislative leadership posts held by Japanese politicians and using a fine-tuned transformer-based machine learning model to classify policy areas in over 46,900 statements from 1270 candidate manifestos across five elections. We find that a higher emphasis on a policy issue increases the probability of securing a legislative post in the same area. This relationship remains consistent across multiple elections and persists even when accounting for candidates' previous legislative leadership roles. We also discover greater congruence in distributive policy areas. Our findings indicate that campaigns provide meaningful signals of policy priorities.

Abstract Do policy priorities that candidates emphasize during election campaigns predict their subsequent legislative activities? We study this question by assembling novel data on legislative leadership posts held by Japanese politicians and using a fine-tuned transformer-based machine learning model to classify policy areas in over 46,900 statements from 1270 candidate manifestos across five elections. We find that a higher emphasis on a policy issue increases the probability of securing a legislative post in the same area. This relationship remains consistent across multiple elections and persists even when accounting for candidates' previous legislative leadership roles. We also discover greater congruence in distributive policy areas. Our findings indicate that campaigns provide meaningful signals of policy priorities.

โ€˜Campaign Communication and Legislative Leadershipโ€™ (with Naofumi Fujimura) appeared in the latest issue of Political Science Research and Methods.

Paper (open access): doi.org/10.1017/psrm...
@psrm.bsky.social

Thread below ๐Ÿงต 1/7

15.07.2025 07:43 ๐Ÿ‘ 22 ๐Ÿ” 4 ๐Ÿ’ฌ 1 ๐Ÿ“Œ 1

๐Ÿ“ฃ Jobs, jobs, jobs!

Aarhus Universityโ€™s Department of Political Science is hiring:

๐Ÿ”น Assistant Professor
๐Ÿ—“ Deadline: 25 August
๐Ÿ”— international.au.dk/about/profil...

๐Ÿ”น Associate Professor
๐Ÿ—“ Deadline: 8 September
๐Ÿ”— international.au.dk/about/profil...

04.06.2025 05:54 ๐Ÿ‘ 102 ๐Ÿ” 102 ๐Ÿ’ฌ 3 ๐Ÿ“Œ 9
Post image

#GESISGuide #AI #LLM
๐Ÿง  New to #prompting LLMs or want to level up your skills?
This guide dives into prompting techniquesโ€”from basics to advanced strategies like reasoning models & persona prompting.
Includes examples, combos, and key limitations to watch out for.

https://t1p.de/yv10f (PDF)

16.06.2025 13:00 ๐Ÿ‘ 10 ๐Ÿ” 5 ๐Ÿ’ฌ 1 ๐Ÿ“Œ 1

Maybe of interest to everyone working in or curious about Computational Social Science (CSS).

If you're looking to build or sharpen your #CSS skills, check out our hands-on courses in the #GESISworkshop program and #GESISfallseminar:

๐Ÿ”— www.gesis.org/workshops
๐Ÿ”— www.gesis.org/fallseminar

18.06.2025 14:29 ๐Ÿ‘ 5 ๐Ÿ” 5 ๐Ÿ’ฌ 0 ๐Ÿ“Œ 0

๐Ÿค Want to get involved?
Present your research, serve as a discussant, or just come as an engaged audience member. Contact us!

04.06.2025 08:48 ๐Ÿ‘ 0 ๐Ÿ” 0 ๐Ÿ’ฌ 0 ๐Ÿ“Œ 0

๐Ÿ“… Seminar Details
๐Ÿ—“๏ธ June 11, 2025
๐Ÿ• 13:00โ€“15:00
๐Ÿ“ Aula 4, Palazzo Hercolani (Strada Maggiore 45, Bologna) & on Microsoft Teams
๐Ÿ“ฉ DM us for the Microsoft Teams link!

04.06.2025 08:48 ๐Ÿ‘ 0 ๐Ÿ” 0 ๐Ÿ’ฌ 1 ๐Ÿ“Œ 0