Sam Wang
@samwang.bsky.social
Neurosciencer, democracy reform studier, plotter of data by hand as first step. At Princeton too. Long-form essays at samwang.substack.com.
And, Donald Trump throws Republicans under the bus. Shutdown ending soon.
November 5, 2025 at 3:40 AM
And, Donald Trump throws Republicans under the bus. Shutdown ending soon.
I found this a table little hard to understand, since it's organized within groups of transportation riders.
If you reweight on estimated shares (56% take public transit), it looks like this:
i.e. 72% of Mamdani supporters use public transit; 49% of Cuomo supporters...and 24% of Sliwa supporters.
If you reweight on estimated shares (56% take public transit), it looks like this:
i.e. 72% of Mamdani supporters use public transit; 49% of Cuomo supporters...and 24% of Sliwa supporters.
November 2, 2025 at 1:22 PM
I found this a table little hard to understand, since it's organized within groups of transportation riders.
If you reweight on estimated shares (56% take public transit), it looks like this:
i.e. 72% of Mamdani supporters use public transit; 49% of Cuomo supporters...and 24% of Sliwa supporters.
If you reweight on estimated shares (56% take public transit), it looks like this:
i.e. 72% of Mamdani supporters use public transit; 49% of Cuomo supporters...and 24% of Sliwa supporters.
Learning my KC geography…on the Kansas side, Joe’s KC Bar-B-Que. barbecue from a working gas station!
October 30, 2025 at 8:18 PM
Learning my KC geography…on the Kansas side, Joe’s KC Bar-B-Que. barbecue from a working gas station!
If you see this, I implore you to post a pic from your device without context.
October 29, 2025 at 6:11 PM
If you see this, I implore you to post a pic from your device without context.
get your sign today! princeton4all.org or write to princeton4all at gmail dot com
October 21, 2025 at 2:31 AM
get your sign today! princeton4all.org or write to princeton4all at gmail dot com
For example, California has 9 Congressional districts that were decided by less than 7 points in 2024.
If voters approve the new map in November, that shrinks to just 1 competitive district statewide, in Riverside/SD counties. It's 40% Hispanic, an ICE target.
samwang.substack.com/p/gerrymande...
If voters approve the new map in November, that shrinks to just 1 competitive district statewide, in Riverside/SD counties. It's 40% Hispanic, an ICE target.
samwang.substack.com/p/gerrymande...
October 19, 2025 at 12:28 PM
For example, California has 9 Congressional districts that were decided by less than 7 points in 2024.
If voters approve the new map in November, that shrinks to just 1 competitive district statewide, in Riverside/SD counties. It's 40% Hispanic, an ICE target.
samwang.substack.com/p/gerrymande...
If voters approve the new map in November, that shrinks to just 1 competitive district statewide, in Riverside/SD counties. It's 40% Hispanic, an ICE target.
samwang.substack.com/p/gerrymande...
Join the movement to make Princeton more affordable and inclusive!
princeton4all.org is a growing movement to help the town grow - and keep on making new history.
Join, and get a free yard sign! Request a sign here: docs.google.com/forms/d/e/1F...
princeton4all.org is a growing movement to help the town grow - and keep on making new history.
Join, and get a free yard sign! Request a sign here: docs.google.com/forms/d/e/1F...
October 18, 2025 at 10:14 PM
Join the movement to make Princeton more affordable and inclusive!
princeton4all.org is a growing movement to help the town grow - and keep on making new history.
Join, and get a free yard sign! Request a sign here: docs.google.com/forms/d/e/1F...
princeton4all.org is a growing movement to help the town grow - and keep on making new history.
Join, and get a free yard sign! Request a sign here: docs.google.com/forms/d/e/1F...
Want to speak up for an affordable, inclusive Princeton? Write to us, and get a yard sign!
princeton4all.org
princeton4all.org
October 18, 2025 at 9:45 PM
Want to speak up for an affordable, inclusive Princeton? Write to us, and get a yard sign!
princeton4all.org
princeton4all.org
"I feel bad for our collapsing Fourth Estate but this is tremendous content" @sallyjenx.bsky.social
October 13, 2025 at 12:13 AM
"I feel bad for our collapsing Fourth Estate but this is tremendous content" @sallyjenx.bsky.social
He's commenting on an essay by Dan Nexon, pointing out that the existing Constitutional order has failed - and the main question is what comes after.
www.lawyersgunsmoneyblog.com/2025/10/the-...
www.lawyersgunsmoneyblog.com/2025/10/the-...
October 7, 2025 at 4:51 PM
He's commenting on an essay by Dan Nexon, pointing out that the existing Constitutional order has failed - and the main question is what comes after.
www.lawyersgunsmoneyblog.com/2025/10/the-...
www.lawyersgunsmoneyblog.com/2025/10/the-...
From Stephen L. Taylor: The Constitution Has Failed. outsidethebeltway.com/has-the-cons...
October 7, 2025 at 4:49 PM
From Stephen L. Taylor: The Constitution Has Failed. outsidethebeltway.com/has-the-cons...
September 27, 2025 at 11:25 PM
Thinking locally: Delighted to show off the new sign for a new campaign for affordability and inclusion.
To learn more, see princeton4all.org
To learn more, see princeton4all.org
September 27, 2025 at 3:19 PM
Thinking locally: Delighted to show off the new sign for a new campaign for affordability and inclusion.
To learn more, see princeton4all.org
To learn more, see princeton4all.org
If anyone wants to see it, a funnel plot of the Baccarelli study looks like this. Basically points toward a minuscule effect, and one can't rule out no effect at all.
September 25, 2025 at 4:06 AM
If anyone wants to see it, a funnel plot of the Baccarelli study looks like this. Basically points toward a minuscule effect, and one can't rule out no effect at all.
...actually did that control. They compared kids with their siblings, thus removing familial confounds.
After that, the risk ratio was 0.98. In other words, a 2% decreased risk (but that's nominal - in fact it's not significant).
And here is their article for you! drive.google.com/file/d/1RWz3...
After that, the risk ratio was 0.98. In other words, a 2% decreased risk (but that's nominal - in fact it's not significant).
And here is their article for you! drive.google.com/file/d/1RWz3...
![To address unobserved confounding, matched full sibling pairs were also analyzed. Sibling control analyses found no evidence that acetaminophen use during pregnancy was associated with autism (HR, 0.98 [95% CI, 0.93-1.04]; RD, 0.02% [95% CI, −0.14% to 0.18%]), ADHD (HR, 0.98 [95% CI, 0.94-1.02]; RD, −0.02% [95% CI, −0.21% to
0.15%]), or intellectual disability (HR, 1.01 [95% CI, 0.92-1.10]; RD, 0% [95% CI, −0.10% to 0.13%]).](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:y4vk2n4vk4ppcderc75uenhh/bafkreic67gttvg3wdfzjcyy25pd3ha3gruxs6of2eoo2fsgjqprbgg4nh4@jpeg)
September 25, 2025 at 3:17 AM
...actually did that control. They compared kids with their siblings, thus removing familial confounds.
After that, the risk ratio was 0.98. In other words, a 2% decreased risk (but that's nominal - in fact it's not significant).
And here is their article for you! drive.google.com/file/d/1RWz3...
After that, the risk ratio was 0.98. In other words, a 2% decreased risk (but that's nominal - in fact it's not significant).
And here is their article for you! drive.google.com/file/d/1RWz3...
Baccarelli admit that they can't rule out that "...there is, as always, a possibility for residual confounding."
For example, an example of a confounding factor would be: what if moms who take Tylenol had a preexisting risk factor that was why they took Tylenol in the first place?
Ahlqvist...
For example, an example of a confounding factor would be: what if moms who take Tylenol had a preexisting risk factor that was why they took Tylenol in the first place?
Ahlqvist...
September 25, 2025 at 3:13 AM
Baccarelli admit that they can't rule out that "...there is, as always, a possibility for residual confounding."
For example, an example of a confounding factor would be: what if moms who take Tylenol had a preexisting risk factor that was why they took Tylenol in the first place?
Ahlqvist...
For example, an example of a confounding factor would be: what if moms who take Tylenol had a preexisting risk factor that was why they took Tylenol in the first place?
Ahlqvist...
Let's look at Dean Baccarelli's best evidence for Tylenol->ASD. Here it is, a meta-analysis of 8 studies.
But they appear to misstate the conclusion of the largest study, of 2.5 *million* kids by Ahlqvist. Baccarelli shows a risk ratio of 1.05 (5% increased risk)... drive.google.com/file/d/1slm2...
But they appear to misstate the conclusion of the largest study, of 2.5 *million* kids by Ahlqvist. Baccarelli shows a risk ratio of 1.05 (5% increased risk)... drive.google.com/file/d/1slm2...
September 25, 2025 at 3:10 AM
Let's look at Dean Baccarelli's best evidence for Tylenol->ASD. Here it is, a meta-analysis of 8 studies.
But they appear to misstate the conclusion of the largest study, of 2.5 *million* kids by Ahlqvist. Baccarelli shows a risk ratio of 1.05 (5% increased risk)... drive.google.com/file/d/1slm2...
But they appear to misstate the conclusion of the largest study, of 2.5 *million* kids by Ahlqvist. Baccarelli shows a risk ratio of 1.05 (5% increased risk)... drive.google.com/file/d/1slm2...
Controlled for things like correlated factors, some % increases in risk. These are all fairly small compared with genetics - but they are things we can control, so we care.
The lowest measurable risk? older parenthood, ratio=1.3 to 1.4. Pretty small.
Guess where Tylenol would be? IT WOULDN'T BE.
The lowest measurable risk? older parenthood, ratio=1.3 to 1.4. Pretty small.
Guess where Tylenol would be? IT WOULDN'T BE.
September 25, 2025 at 3:03 AM
Controlled for things like correlated factors, some % increases in risk. These are all fairly small compared with genetics - but they are things we can control, so we care.
The lowest measurable risk? older parenthood, ratio=1.3 to 1.4. Pretty small.
Guess where Tylenol would be? IT WOULDN'T BE.
The lowest measurable risk? older parenthood, ratio=1.3 to 1.4. Pretty small.
Guess where Tylenol would be? IT WOULDN'T BE.
Here's the key chart from that piece.
Note what's on here with significant risk: genetic inheritance (twin on autism spectrum, sibling), third trimester stress, and factors that can affect maternal/paternal DNA.
What doesn't carry risk? SSRIs. Sonograms. And vaccination. All very well tested.
Note what's on here with significant risk: genetic inheritance (twin on autism spectrum, sibling), third trimester stress, and factors that can affect maternal/paternal DNA.
What doesn't carry risk? SSRIs. Sonograms. And vaccination. All very well tested.
September 25, 2025 at 2:59 AM
Here's the key chart from that piece.
Note what's on here with significant risk: genetic inheritance (twin on autism spectrum, sibling), third trimester stress, and factors that can affect maternal/paternal DNA.
What doesn't carry risk? SSRIs. Sonograms. And vaccination. All very well tested.
Note what's on here with significant risk: genetic inheritance (twin on autism spectrum, sibling), third trimester stress, and factors that can affect maternal/paternal DNA.
What doesn't carry risk? SSRIs. Sonograms. And vaccination. All very well tested.
Some years ago, I wrote that the odds ratio can help us put a risk we read about into perspective. www.nytimes.com/2014/03/30/o...
Take the rate in the general population, and then see what the *true* added risk is from adding a likely-causative factor. This puts a lot of work into perspective.
Take the rate in the general population, and then see what the *true* added risk is from adding a likely-causative factor. This puts a lot of work into perspective.
September 25, 2025 at 2:57 AM
Some years ago, I wrote that the odds ratio can help us put a risk we read about into perspective. www.nytimes.com/2014/03/30/o...
Take the rate in the general population, and then see what the *true* added risk is from adding a likely-causative factor. This puts a lot of work into perspective.
Take the rate in the general population, and then see what the *true* added risk is from adding a likely-causative factor. This puts a lot of work into perspective.