Aaron Gullickson

What students think working with data looks like. #EconSky

There is no right answer to what constitutes a "generation" but there are definitely wrong ones!

As some of you guessed, the two models are statistically indistinguishable. The answer is Model 1, but there was no way to tell from the data. I was very surprised when I first saw this. Here we show that this confounding of heterogeneity and reinforcement is general: www.pnas.org/doi/10.1073/...

I have heard this claim a couple of places now and it is idiotic. Do you really think Iran was somehow unaware of the value of nuclear deterrence until this attack? Somehow, Iran missed a fact that has been known since the beginning of the Cold War? They were just enriching uranium for kicks?

❤️ @migration.ubc.ca for featuring my research! My project on immigrants' relationship formation is driven by my belief that calling Canada home isn’t just about work, but also about love & connection. Supporting immigrants in building meaningful relationships fosters well-being, belonging & thriving

Check all your group chats for bombing plans

Nice! I did not know it had built in bootstrapping functionality.

bootstap_me_inequality.R GitHub Gist: instantly share code, notes, and snippets.

Ok, here is a simple repex of how you could go about bootstrapping the ME inequality measure.

gist.github.com/AaronGullick...

If you need to bootstrap, I would still use avg_comparisons() to get your ME inequality estimate, but just set vcov = FALSE so it runs faster and you don't need the SE. Then just for loop through B times, resample data with replacement, and combine estimates from which you can draw 95% CI and SE.

Thanks, this is a really interesting article. I just skimmed it but I did not see a suggestion that the delta method wouldn't work well, except for maybe taking a bit longer with many categories. Have you tried marginaleffects::avg_comparisons and found it doesn't work well?

Not sure I follow. Do you mean a literal bootstrap approach? If so, you don't need the computed SEs. If not, are the values from a model? I have found marginaleffects can handle many such cases w the right model setup.

Democrats’ losses among Hispanic voters may be larger than they look. Conservative Hispanic Americans are ditching their Hispanic identity, surveys show.

it's not just that Hispanic voters shifted Republican since 2020, it's also the case that some of them no longer even identify as Hispanic! new @goodauth.bsky.social post from me and Caroline Soler
goodauthority.org/news/democra...

Pitfalls of premature closure with LLM assisted coding When LLM models generates clean, professional-looking code, it's tempting to stop exploring alternatives. But therein lies the risks that comes with premature closure. So what is premature closure?

New to me is the term "premature closure", where you too quickly latch on to the first solution you see. Always a danger in coding, but particularly so today when LLMs can give you a plausible fix so so quickly.

www.shayon.dev/post/2025/16...

Arrested Development meme. People have suggested empirically solving the APC problem just by applying the right model to the right data Well did it work for those people Not, it never does, I mean, these people somehow delude themselves into thinking it might, but... ...but it might work for us.

PSA: Don’t trust anyone who tells you that you can identify age or period or cohort effects simply by applying the right statistical model to the right type of data. This is fundamentally misunderstanding the nature of the age-period-cohort problem!>

In Memoriam: Professor Emeritus Richard Alba An eminent demographer and sociologist, he challenged conventional thinking on how immigrants integrate into U.S. society.

The Graduate Center community deeply mourns the passing of Prof. Emeritus Richard Alba, a pathbreaking demographer who bucked conventional ideas about how immigrants assimilate into and impact mainstream U.S. culture www.gc.cuny.edu/news/memoria...

I can't handle all the various academic takes on effective protests right now, especially all the usual callbacks to the Civil Rights Movement. Too many people know exactly the right tactical approach here. The correct answer is you don't know; do what you think is right.

I can't believe I never had the insight to add clip art of famous architectural features to my plots! It really accelerates the absorption of critical insights.

Jessica Vasquez-Tokos, Podcast Episode · New Books in Sociology · 06/02/2025 · 33m

My first episode as a host for @newbooksnetwork.bsky.social is out! It was truly a joy to discuss Burdens of Belonging with Jessica Vasquez-Tokos! We talked Du Bois, moving beyond the black white binary, methods, and the power of stories and history! Check it out : podcasts.apple.com/us/podcast/n...

GitHub - jogaudard/phdthesis_toolbox: Some technical tricks to write a PhD thesis in Quarto Some technical tricks to write a PhD thesis in Quarto - jogaudard/phdthesis_toolbox

#phd thesis at @unibergen.bsky.social have to include a table with co authors of included articles, contributions and affiliations.
Here is a @github.com template to do this in #quarto, using the Plume R package and multibib filter.
@btwnthefjords-uib.bsky.social @bjerknes.uib.no #rstats #phdLife

Article title: Agenda Seeding: How 1960s Black Protests Moved Elites, Public Opinion and Voting Article Abstract How do stigmatized minorities advance agendas when confronted with hostile majorities? Elite theories of influence posit marginal groups exert little power. I propose the concept of agenda seeding to describe how activists use methods like disruption to capture the attention of media and overcome political asymmetries. Further, I hypothesize protest tactics influence how news organizations frame demands. Evaluating black-led protests between 1960 and 1972, I find nonviolent activism, particularly when met with state or vigilante repression, drove media coverage, framing, congressional speech, and public opinion on civil rights. Counties proximate to nonviolent protests saw presidential Democratic vote share increase 1.6–2.5%. Protester-initiated violence, by contrast, helped move news agendas, frames, elite discourse, and public concern toward “social control.” In 1968, using rainfall as an instrument, I find violent protests likely caused a 1.5–7.9% shift among whites toward Republicans and tipped the election. Elites may dominate political communication but hold no monopoly.

I’ve been studying civil rights protests for 20 years. With new mobilization against Trump’s agenda, I’m sharing a thread summarizing my research on how nonviolent & violent actions by 1960s activists and police influenced media, elites, public opinion & voters. 1/ www.cambridge.org/core/journal...

Really it's just a repackaging of the same concerns and weaknesses of observational data especially in the cross-section. You just don't know which way the arrow is pointing.

This is the first term of teaching stats where I am really seeing a lot of LLM use and the use of %>% is the key tipoff given that I never showed students this. They are also doing stuff like group_by |> group_modify when I showed them group_by |> group_split |> map. At least I am learning things!

Not my best data viz, but this is my favorite exercise in undergrad stats class. I ask students to guess what the gender gap is in abortion support and then we all calculate it in class. Its an excellent demonstration of why good data analysis matters, and we shouldn't believe everything we think.

Please just block me if you think setting Jewish civilians on fire is good praxis. Save us both the trouble.

Spillover or consolidation?

We tested 2 models of cultural polarization w/ ~2,000 Americans.

The data support consolidation: once we accounted for networks and dispositions, partisanship lost its explanatory power.

📄 Childress, Rawlings, & Maghbouleh (2025)

🔗 doi.org/10.31235/osf.io/zpe8y_v

Changes in the College Mobility Pipeline Since 1900 Founded in 1920, the NBER is a private, non-profit, non-partisan organization dedicated to conducting economic research and to disseminating research findings among academics, public policy makers, an...

Important paper by Bleemer and Quincy: Since 1960, the college wage premium has become less equal. Lower-income students now get far less out of college than their higher-income peers.

www.nber.org/papers/w33797

If valedictorian addresses had a strong tradition of being apolitical, that would be one thing. But they're routinely characterized by conventional center-left political pieties presented as obviously shared, we-all-love-apple-pie, aspirations for the future. And that's OK!
1/

Interesting approach, this is