Dan de Kadt
@dandekadt.bsky.social
Social and data science at the London School of Economic
Trying to come up with a good idea
Trying to come up with a good idea
Pinned
Dan de Kadt
@dandekadt.bsky.social
· May 14
good_description/good_description_ddk_agb.pdf at main · ddekadt/good_description
Homepage of "Good Description" by Daniel de Kadt & Anna Grzymala-Busse - ddekadt/good_description
tinyurl.com
🚨 “Good Description” with @annagbusse.bsky.social 🚨
What sets 'good' description apart from 'mere' description?
We develop a framework for evaluating descriptive research, whether we are doing it as scholars or assessing it as readers.
Two main contributions...
🔗📄 tinyurl.com/gooddesc
What sets 'good' description apart from 'mere' description?
We develop a framework for evaluating descriptive research, whether we are doing it as scholars or assessing it as readers.
Two main contributions...
🔗📄 tinyurl.com/gooddesc
Reposted by Dan de Kadt
The LSE finding itself on this list of woke American schools like
February 14, 2026 at 7:23 AM
The LSE finding itself on this list of woke American schools like
I’m tired Claude
February 13, 2026 at 11:23 PM
I’m tired Claude
Reposted by Dan de Kadt
DHS is shutting down because it refuses to obey the law. They are tear-gassing schools, killing American citizens, and disappearing legal immigrants. Democrats shouldn’t fund an out of control ICE.
February 13, 2026 at 4:53 PM
DHS is shutting down because it refuses to obey the law. They are tear-gassing schools, killing American citizens, and disappearing legal immigrants. Democrats shouldn’t fund an out of control ICE.
Irony being each one of us is actually just measurement error to BLS
February 13, 2026 at 10:36 PM
Irony being each one of us is actually just measurement error to BLS
Forsooth I shall become Caesar by speaking like Caesar
“I have searched the depths of Legal Arguments not yet articulated or vetted on this subject, and will be presenting an irrefutable one in the very near future”
February 13, 2026 at 10:27 PM
Forsooth I shall become Caesar by speaking like Caesar
I was submitting slop to journals before AI made it cool 😎
February 13, 2026 at 12:21 PM
I was submitting slop to journals before AI made it cool 😎
Whether you agree with all of the below or not, Berna is imo the person doing the most interesting thinking in the metascience space. It’s provocative (cuts against the accepted wisdom) but very important.
I do really strongly agree with her about the crisis/disaster narrative.
I do really strongly agree with her about the crisis/disaster narrative.
On the publication bias discourse, I regret that metascience has become a source of decontextualized, low-res, bean-counting-focused `science is in crisis' narratives. It is largely uncurious abt science, desperately lacking in theory & measurement. I'll quote a few takes I liked & add my thoughts🧵
February 13, 2026 at 8:43 AM
Whether you agree with all of the below or not, Berna is imo the person doing the most interesting thinking in the metascience space. It’s provocative (cuts against the accepted wisdom) but very important.
I do really strongly agree with her about the crisis/disaster narrative.
I do really strongly agree with her about the crisis/disaster narrative.
Reposted by Dan de Kadt
We invite nominations for the @apsa.bsky.social CP Section's Margaret Levi Award for the Advancement of Comparative Methodology. Self-nominations welcome! Deadline: March 15, 2026. Send nominations to golden@ucla.edu. Details 👇
February 13, 2026 at 2:39 AM
We invite nominations for the @apsa.bsky.social CP Section's Margaret Levi Award for the Advancement of Comparative Methodology. Self-nominations welcome! Deadline: March 15, 2026. Send nominations to golden@ucla.edu. Details 👇
Just for clarity, and it’s a small detail of course but important, but two people are actually dead because federal agents pulled guns, pointed the guns at them, and then pulled the trigger multiple times.
(We have the recordings.)
(We have the recordings.)
Ron Johnson to Keith Ellison: "Two people are dead because you encouraged them to put themselves into harm's way! And now you are exploiting those two martyrs. You ought to feel damn guilty about it. Yeah, sit there and smirk. Smirk. It's sick! It's despicable. You disgust me."
February 12, 2026 at 7:12 PM
Just for clarity, and it’s a small detail of course but important, but two people are actually dead because federal agents pulled guns, pointed the guns at them, and then pulled the trigger multiple times.
(We have the recordings.)
(We have the recordings.)
Reposted by Dan de Kadt
Ending this operation is not enough.
We need justice and accountability. That starts with independent investigations into the murders of Renee Good and Alex Pretti, economic restitution for businesses impacted, abolishing ICE, and the impeachment of Kristi Noem.
We need justice and accountability. That starts with independent investigations into the murders of Renee Good and Alex Pretti, economic restitution for businesses impacted, abolishing ICE, and the impeachment of Kristi Noem.
February 12, 2026 at 5:07 PM
Ending this operation is not enough.
We need justice and accountability. That starts with independent investigations into the murders of Renee Good and Alex Pretti, economic restitution for businesses impacted, abolishing ICE, and the impeachment of Kristi Noem.
We need justice and accountability. That starts with independent investigations into the murders of Renee Good and Alex Pretti, economic restitution for businesses impacted, abolishing ICE, and the impeachment of Kristi Noem.
Surely the UK government will follow suit
Today, the Government of Ontario announced changes to funding across the postsecondary sector, including a $6.4 billion investment.
This funding will help sustain access to high-quality academic programs & support students in preparing for the future. uoft.me/funding-announcement
This funding will help sustain access to high-quality academic programs & support students in preparing for the future. uoft.me/funding-announcement
February 12, 2026 at 5:25 PM
Surely the UK government will follow suit
Reposted by Dan de Kadt
Doing my share to reduce the bias of the published research record by file drawering a study with significant results
I have a new paper. We look at ~all stats articles in political science post-2010 & show that 94% have abstracts that claim to reject a null. Only 2% present only null results. This is hard to explain unless the research process has a filter that only lets rejections through.
February 12, 2026 at 1:47 PM
Doing my share to reduce the bias of the published research record by file drawering a study with significant results
What's bigger here, the lie or the collar gap?
Homan: "ICE is a legitimate federal law enforcement agency. We're not out scouring the streets to disappear people or deny people their civil rights or due process."
February 12, 2026 at 4:15 PM
What's bigger here, the lie or the collar gap?
And into the causal inference slide deck you go!
Looking for an edge in college admissions? Just pull your kid from school. Homeschooling, long associated with hippies and religious conservatives in the U.S., is in the middle of a rebrand and a boom.
The Homeschooling Hack
Looking for an edge in college admissions? Just pull your kid from school. Homeschooling, long associated with hippies and religious conservatives in the U.S., is in the middle of a rebrand and a boom.
nymag.com
February 12, 2026 at 2:37 PM
And into the causal inference slide deck you go!
You: playing Russian roulette in the basement of an illegal casino
Me: playing with my son's whoopie cushion while on a zoom call
We are not the same
Me: playing with my son's whoopie cushion while on a zoom call
We are not the same
February 12, 2026 at 2:24 PM
You: playing Russian roulette in the basement of an illegal casino
Me: playing with my son's whoopie cushion while on a zoom call
We are not the same
Me: playing with my son's whoopie cushion while on a zoom call
We are not the same
Reposted by Dan de Kadt
When R2 asks you to do those robustness checks you really don't want to do, just replying "The Dow is over $50,000!" is always an option.
February 12, 2026 at 3:01 AM
When R2 asks you to do those robustness checks you really don't want to do, just replying "The Dow is over $50,000!" is always an option.
Reposted by Dan de Kadt
i'd like to see AI do this!!! [draws a convex function which i label concave and discuss as if it's concave for the entire lecture despite the vague feeling that i've made some kind of error, only realising next week what happened]
AI simply cannot stand in front of a classroom, gesticulating wildly and writing semi-illegibly on chalkboards, both of which are key to building rapport and convincing your students you are worth learning from
If people think AI can replace historians and political scientists, I have a bridge on the Thames to sell you.
February 12, 2026 at 2:55 AM
i'd like to see AI do this!!! [draws a convex function which i label concave and discuss as if it's concave for the entire lecture despite the vague feeling that i've made some kind of error, only realising next week what happened]
Whew, my Claude has a very low pr(dementia)
February 12, 2026 at 10:24 AM
Whew, my Claude has a very low pr(dementia)
I a very grateful to Ryan/Vincent/Jon for taking on this work, and I absolutely agree with (1), but I still think Ryan is too pessimistic about (2).
Why? I think if you look at the last ~50 years of empirical social science, we've learned a *lot* about how the world works, and how people work.
Why? I think if you look at the last ~50 years of empirical social science, we've learned a *lot* about how the world works, and how people work.
Overall, I think our results are:
1. Believable. I think deep down we all know we publish "too many" significant results.
2. Disastrous. Selecting on significance in this way biases results across our whole literature away from zero & it stops us from learning what doesn't work.
1. Believable. I think deep down we all know we publish "too many" significant results.
2. Disastrous. Selecting on significance in this way biases results across our whole literature away from zero & it stops us from learning what doesn't work.
February 12, 2026 at 9:09 AM
I a very grateful to Ryan/Vincent/Jon for taking on this work, and I absolutely agree with (1), but I still think Ryan is too pessimistic about (2).
Why? I think if you look at the last ~50 years of empirical social science, we've learned a *lot* about how the world works, and how people work.
Why? I think if you look at the last ~50 years of empirical social science, we've learned a *lot* about how the world works, and how people work.
Reposted by Dan de Kadt
Monsters. Absolute monsters. From the front line to the Oval Office.
More than 400 people, including hundreds of citizens, were “sorted” at gun point into racial and ethnic groups by 200 ICE agents who fired flash-bang grenades into cars with people inside, pointed guns at children & demanded their zip tied parents not comfort them. apple.news/Ausdx6kWtT0e...
Idaho families sue over immigration raid that swept up hundreds, including U.S. citizens — NBC News
About 400 people, including children and U.S. citizens, were detained for four hours while they were denied food and water in the raid, according to the lawsuit.
apple.news
February 12, 2026 at 1:57 AM
Monsters. Absolute monsters. From the front line to the Oval Office.
Way too much of our reduced form causal inference work follows from “it is a truth universally acknowledged that a p-value under 0.05 should be in want of a theory”
It's ironic to see a discipline care **so much** about unbiasedness (causal inference!) at the level of a single test but then have a research production system and culture that is basically a ferocious bias generation machine. This is not good.
February 11, 2026 at 9:19 PM
Way too much of our reduced form causal inference work follows from “it is a truth universally acknowledged that a p-value under 0.05 should be in want of a theory”
Reposted by Dan de Kadt
ICE are still being terrible in the Twin Cities, hurting people and causing mayhem.
This is the aftermath of an ICE kidnapping a few blocks from my home in St. Paul—an hour ago. A quiet street full of broken glass and at least three wrecked cars. The target of the kidnapping was taken away by ambulance. He was on a stretcher and covered by a sheet, though a cop said he was alive.
February 11, 2026 at 8:21 PM
ICE are still being terrible in the Twin Cities, hurting people and causing mayhem.
Y’all should just do more good description.
Where p-values don’t even (have to) matter!*
*jokes, p-values never have to matter
Where p-values don’t even (have to) matter!*
*jokes, p-values never have to matter
When I pitch academics on my paper on nulls one common and understandable reaction is "but they're probably noisy and thus uninformative nulls." This is true, but it misses the key realization that WE PUBLISH THE RESULT WHEN THE NOISY TEST IS P<0.05.
February 11, 2026 at 8:21 PM
Y’all should just do more good description.
Where p-values don’t even (have to) matter!*
*jokes, p-values never have to matter
Where p-values don’t even (have to) matter!*
*jokes, p-values never have to matter
In before everyone says “proudly the 2%”
(Proudly the 2%)
(Proudly the 2%)
I have a new paper. We look at ~all stats articles in political science post-2010 & show that 94% have abstracts that claim to reject a null. Only 2% present only null results. This is hard to explain unless the research process has a filter that only lets rejections through.
February 11, 2026 at 7:23 PM
In before everyone says “proudly the 2%”
(Proudly the 2%)
(Proudly the 2%)
Reposted by Dan de Kadt
"Silence is not neutrality; it is permission."
Wise words on what higher education needs right now from
@arneduncan.bsky.social
www.washingtonpost.com/opinions/202...
Wise words on what higher education needs right now from
@arneduncan.bsky.social
www.washingtonpost.com/opinions/202...
www.washingtonpost.com
February 10, 2026 at 1:59 PM
"Silence is not neutrality; it is permission."
Wise words on what higher education needs right now from
@arneduncan.bsky.social
www.washingtonpost.com/opinions/202...
Wise words on what higher education needs right now from
@arneduncan.bsky.social
www.washingtonpost.com/opinions/202...