Dean Eckles

Research Intern - Computational Social Science | Microsoft Careers Research Interns put inquiry and theory into practice. Alongside fellow doctoral candidates and some of the world's best researchers, Research Interns learn, collaborate, and network for life. Researc...

Want to be an intern at Microsoft Research in the Computational Social Science group in NYC (Jake Hofman, David Rothschild, Dan Goldstein)

Follow this link and do your thing! Deadline approaching soonish!

apply.careers.microsoft.com/careers/job/...

Do these models allow for any decreases?

The perplexing “connected cluster axiom” – Inverse Complexity Lab Research group on inverse problems in complex systems and network science.

I wrote a blog post about the often stated but never explained assumption that communities in graphs should always be connected.

This is inconsistent with statistical significance and null models that underlie the most widely employed methods.

skewed.de/lab/posts/co...

I can see this resonating with my colleagues who have been doing sophisticated conjoint experiments for decades...

We should probably be placing incarcerated people in prisons closest to their homes.

Why?

Assigning individuals to prisons closer to their home reduces recidivism.

Really?

Yep. Being placed close to one's home increases social contacts that appear to reduce reoffending.

🚨 New in Nature+Science!🚨
AI chatbots can shift voter attitudes on candidates & policies, often by 10+pp
🔹Exps in US Canada Poland & UK
🔹More “facts”→more persuasion (not psych tricks)
🔹Increasing persuasiveness reduces "fact" accuracy
🔹Right-leaning bots=more inaccurate

They’re killing people as a twitter joke. It’s depraved. Can’t let yourself lose the capacity to be appalled and disgusted and outraged by this.

helps to think about the details like any regular crime you'd see on one of the 800 cop shows that run on TV constantly: these guys committed one set of murders, thought it over, and then decided to kill the witnesses as part of the coverup. mention of drugs is just jingling keys at this point

A blog post giving a more thorough take on survey experiments and the credibility revolution: cyrussamii.com?p=4168

sex1 sex2 sex3 sex4 sex5 sex6 sex7 sex8 Boy Girl Boy Boy Boy Boy Girl Girl Boy Girl Girl Boy Boy Boy Girl Boy Girl Girl Boy Girl Boy Boy Boy Girl Boy Boy Boy Girl Boy . Boy . Girl Boy Boy Boy Girl Girl Girl Girl Boy . . . . Boy

What if every family stops having children as soon as they have a boy? Would the sex ratio end up skewed one way or the other?

This is what it would look like for 25 random families following that rule (with a normal male/female sex ratio of 105/100.)

/1

You’re absolutely right — you are Pagliacci. It would certainly be difficult for you to attend your own performance! I should not have given such paradoxical advice, and I apologize deeply for the error. There is no excuse for my failure.

Haidt is "stunned" that ChatGPT asked how it would threaten youth reveals that its plan would be what Haidt has argued in his books. The secret plan!

There should be some way to rescind the doctorate of any commentator on digital media who interprets ChatGPT output like this. I am not joking.

This is amazing. If you haven't reveled in the glory that is the IPUMS international census archive, you should! Here's something I made from it: "The Rise of One-Person Households" journals.sagepub.com/doi/full/10....

I think they may not actually be statistically distinguishable. But not sure... Everything in the FBES papers is SD units so I'm not sure what the SE for affective polarization in points is.

We first note that hand-wringing about the decline in US college enrollments has mistakenly linked such declines to the price of four-year colleges.

But the decline is entirely driven by two-year community colleges (and by for-profit colleges). The four-year sector is the dog that didn't bark.

What would you do if your 2SLS estimates are much larger than OLS, but you still have confidence in the instrument? Any tools or techniques to further probe what's going on?

In that case you could try weightIng the OLS with the implicit TSLS weights. Or weight the TSLS with inverse complier scoring weights. So then if there is selection on observables etc, they should agree.

Today, social media platforms hold the sole power to study the effects of feed-ranking algorithms. We developed a platform-independent method that reranks participants’ feeds in real time and used this method to conduct a preregistered 10-day field experiment with 1256 participants on X during the 2024 US presidential campaign. Our experiment used a large language model to rerank posts that expressed antidemocratic attitudes and partisan animosity (AAPA). Decreasing or increasing AAPA exposure shifted out-party partisan animosity by more than 2 points on a 100-point feeling thermometer, with no detectable differences across party lines, providing causal evidence that exposure to AAPA content alters affective polarization. This work establishes a method to study feed algorithms without requiring platform cooperation, enabling independent evaluation of ranking interventions in naturalistic settings.

New paper in Science:

In a platform-independent field experiment, we show that reranking content expressing antidemocratic attitudes and partisan animosity in social media feeds alters affective polarization.

🧵

For the avoidance of doubt: does svyglm() use robust standard errors? Yes!
notstatschat.rbind.io/2025/12/01/d...

Zillow Removes Climate Risk Scores From Home Listings

You can still find property level climate risk scores on Redfin

Zillow Removes Climate Risk Scores From Home Listings
www.nytimes.com/2025/11/30/c...

I think it computes SEs in a way that is justified by the design, but which also turns out to largely coincide with robust SEs

They’re burying the weirdest part of this story, which is that this congressman’s twin brother keeps trying and failing to replace him in various jobs and is now hoping to replace him in this one

Politics Is Seeping Into Our Daily Life and Ruining Everything Is there anything that politics can't ruin? The answer, it appears, is a resounding "no" as partisan conflict creeps into…

Makes me think of the "oil spill" model of polarization, and broader discussion by @jdtuccille.bsky.social here
reason.com/2021/02/17/p...

Does S.F. have the ‘infrastructure’ to support Lurie’s housing plan? Let’s consider Housing opponents love to say we lack “the infrastructure” to accommodate new neighbors our otherwise inclusive communities. What are they talking about?

The “we can’t possibly build more homes because our infrastructure isn’t perfect” argument is ubiquitous in San Francisco housing debates.

It’s effectively saying “sorry, no home for you until all infrastructure is perfect — go live 100 miles away.”

The argument is also factually baseless:

The surprising crab-trap–pulling behavior is described in this article: onlinelibrary.wiley.com/doi/10.1002/...

We recently published a paper on how animal tool use is conceptualized, offering a framework to make sense of its different forms 👇📃 www.cambridge.org/core/journal... #evosky #philsci

We shouldn’t be too quick to see inconsistent beliefs as evidence of irrationality.

Cool research by @bayesandbounds.bsky.social & @tanialombrozo.bsky.social suggests inaccessibility of an apparently conflicting belief can lead to (temporary, quickly resolved) inconsistency:

buff.ly/QncPtjq

Forecasting Social Science: Evidence from 100 Projects 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...

"Confidence in the accuracy of one's forecasts is perversely associated with lower accuracy."

www.nber.org/papers/w34493

Satellite image of a commercial development along a stroad. On the left there's a big box store that says "Property taxes: $102,485, Parcel Size: 17,766 meter square. On the right there's two apartment buildings that read "Property taxes: $626,002, Parcel Size: 10,930 meter square"

Minimum parking requirements create financially insolvent land use patterns.

The two apartment buildings on the right generate six times more in property taxes than the big box store on the left, while occupying almost half the space!

#BlackFridayParking

Pete Hegseth is Secretary of Defense *because* he championed war criminals. Literal war criminals convicted of heinous offenses in US courts martial. Got Trump to pardon some in term 1.

Not in spite of, because.

That's what Trump appointed, and what 50 GOP Senators (plus Vance) voted to confirm.

Politics Is Seeping Into Our Daily Life and Ruining Everything Is there anything that politics can't ruin? The answer, it appears, is a resounding "no" as partisan conflict creeps into…

Makes me think of the "oil spill" model of polarization, and broader discussion by @jdtuccille.bsky.social here
reason.com/2021/02/17/p...