Stefan McCabe

counterpoint: don’t visualize networks

ryan’s substack era

DomainDemo: a dataset of domain-sharing activities among different demographic groups on Twitter - Scientific Data Scientific Data - DomainDemo: a dataset of domain-sharing activities among different demographic groups on Twitter

ICYMI, our DomainDemo dataset, which describes how different demographic groups share domains on Twitter, is now available to download!

📄 Data descriptor: doi.org/10.1038/s415...
📈 Interactive app to explore the data: domaindemo.info
💽 Dataset: doi.org/10.5281/zeno...

Using co-sharing to identify use of mainstream news for promoting potentially misleading narratives - Nature Human Behaviour Goel et al. examine why some factually correct news articles are often shared by users who also shared fake news articles on social media.

In our new paper (w/ @jongreen.bsky.social , @davidlazer.bsky.social, & Philip Resnik), now up in Nature Human Behaviour (nature.com/articles/s41562-025-02223-4), we argue that this tension really speaks to a broader misconceptualization of what misinformation is and how it works.

A passage from VS Naipaul’s A House for Mr Biswas: And that was what Mr Biswas continued to feel about their venture: that it was temporary and not quite real, and it didn't matter how it was arranged. He had felt that on the first after-noon; and the feeling lasted until he left The Chase. Real life was to begin for them soon, and elsewhere. The Chase was a pause, a preparation.

A Postdoc for Mr Biswas

If you study networks, or have been stuck listening to people who study networks for long enough (sorry to my loved ones), you may have heard that open triads – V shapes – in social networks tend to turn into closed triangles. But why does this happen? In part, because people repost each other.

the best slack emoji, kool-aid oh yeah

Sage Journals: Discover world-class research Subscription and open access journals from Sage, the world's leading independent academic publisher.

have you ever lost sleep wondering how hashtag activism campaigns unfold over time???

in a new paper with @erikavmelder.bsky.social and @foucaultwelles.bsky.social, we look at the temporal evolution of 7 years of #StopLine3 on twitter

more here: doi.org/10.1177/2056...

Our open-access article of the week shows how individual users—not just algorithms—selectively share stories that align with their identities and values, filtering content into ideological silos.

By @jongreen.bsky.social, @sdmccabe.com et al. in @apsrjournal.bsky.social

https://buff.ly/41m9Thk

congrats!!

congrats! finally!

ten years ago: Twitter Is Real Life (😃)
five years ago: Twitter Is Not Real Life
now: Twitter Is Real Life (😱)

it’s punishment for years of cops at pride discourse

i think we wrote some of the initial code when trump was last in office so kind of a full circle moment

In the 10 years since last week, I neglected to share that I’m so excited that our “Curation Bubbles” paper with @jongreen.bsky.social, @sdmccabe.com, @davidlazer.bsky.social and others is now out in APSR!

if everything is a bipartite network, nothing is a bipartite network

Title: Curation Bubbles Abstract: Information on social media is characterized by networked curation processes in which users select other users from whom to receive information, and those users in turn share information that promotes their identities and interests. We argue this allows for partisan “curation bubbles” of users who share and consume content with consistent appeal drawn from a variety of sources. Yet, research concerning the extent of filter bubbles, echo chambers, or other forms of politically segregated information consumption typically conceptualizes information’s partisan valence at the source level as opposed to the story level. This can lead domain-level measures of audience partisanship to mischaracterize the partisan appeal of sources’ constituent stories—especially for sources estimated to be more moderate. Accounting for networked curation aligns theory and measurement of political information consumption on social media.

Figure 1: Stylized Examples a) Users consuming information directly from sources b) Users curating information for other users

Figure 4: URL Scores by Share Volume for Selected Domains on Twitter and Facebook

Figure 8: Proportion of URLs Substantively Distinct from Domain for Different Facebook Engagement Types

Extremely happy to share that "Curation Bubbles" is online (open access!) at @apsrjournal.bsky.social: www.cambridge.org/core/journal...

Introducing “DomainDemo: a dataset of domain-sharing activities among different demographic groups on Twitter.”

Today, we release five derived metrics of over 129,000 domains, quantifying their characteristics such as geographical reach and audience partisanship.

1/3

it’s 2025, we’re doing everything in SQL now

duckpgq DuckDB Community Extensions Extension that adds support for SQL/PGQ and graph algorithms

duckdb.org/community_ex...

Kublai Khan in Invisible Cities is the original large language model

you can tell if someone gets their bibtex from google scholar if it’s terrible bibtex

Congratulations!!

Democratizing Algorithmic Feeds on Bluesky If the original sin of Web 1.0 was the pop-up ad, the original sin of web 2.0 was the move to algorithmic feeds. Opaque optimization strategies aimed at maximizing private revenue for the sake of what...

Over the past few days I've been working out some thoughts about how the Bluesky feed generator feature could be used to, just maybe, make the internet a healthier place this time around. I think I have a sketch of a path there, and I'd love to talk it through! devingaffney.com/democratizin...

i can denounce you in the quote tweets if you’d like

if you want a break from the election, we've posted a revision of our paper on pundits and ideological coalitions. bigger emphasis in this draft on the point that political ideologies aren't quite the same thing as political philosophies osf.io/xfy8r

debugging python environments builds character - if you aren’t calling `which` dozens of times a day you aren’t living

Very honored to share that "Inequalities in Online Representation: Who Follows Their Own Member of Congress on Twitter?" (w/ @sdmccabe.com, Pranav Goel, and @davidlazer.bsky.social) won the 2024 Best Article award from APSA's Information Technology and Politics section! journalqd.org/article/view...

Post-January 6th deplatforming reduced the reach of misinformation on Twitter - Nature Difference-in-differences analysis indicates that the decision by Twitter to deplatform 70,000 users following the events at the US Capitol on 6 January 2021 had wider effects on the spread of mi...

New paper in Nature looking at the direct *and spillover* effects of Twitter deplatforming ~70k accounts following January 6th:
- less misinfo for remaining users who had followed deplatformed users to circulate
- some remaining users voluntarily exited
www.nature.com/articles/s41...

Title: Curation Bubbles Abstract: Information on social media is characterized by networked curation processes in which users select other users from whom to receive information, and those users in turn share information that promotes their identities and interests. We argue this allows for partisan "curation bubbles" of users who share and consume content with consistent appeal drawn from a variety of sources. Yet, research concerning the extent of filter bubbles, echo chambers, or other forms of politically segregated information consumption typically conceptualizes information's partisan valence at the source level as opposed to the story level. This can lead domain-level measures of audience partisanship to mischaracterize the partisan appeal of sources' constituent stories -- especially for sources estimated to be more moderate. Accounting for networked curation aligns theory and measurement of political information on social media.

Figure 1: Stylized examples of direct consumption and partisan curation, with curators mediating information flows in the latter.

Figure 4b: URL audience scores by share volume for selected domains on Facebook, showing proportions of URLs for each domain that are statistically/substantively distinguishable from their parent domain's overall audience score.

Figure 7: Proportion of URLs substantively distinct from domain by domain-level audience score. Domain-level audience scores closer to zero more frequently mischaracterize the audience scores of those domains' constituent stories.

Very happy to share that "Curation Bubbles" is conditionally accepted at the American Political Science Review. osf.io/vbwer