New interactive blog!
"Why Adjusted Regression Coefficients Are Less Descriptive Than They Look"
rpsychologist.com/descriptive-...
"Why Adjusted Regression Coefficients Are Less Descriptive Than They Look"
rpsychologist.com/descriptive-...
June 25, 2026 at 10:59 AM
New interactive blog!
"Why Adjusted Regression Coefficients Are Less Descriptive Than They Look"
rpsychologist.com/descriptive-...
"Why Adjusted Regression Coefficients Are Less Descriptive Than They Look"
rpsychologist.com/descriptive-...
One thing that just keeps giving is @vincentab.bsky.social's marginaleffects package (see marginaleffects.com). It's just a complete game changer, moving from "trying to discern what model coefficients tell you after careful recoding" to "you know what, just compare these slopes, thx"
December 13, 2024 at 12:01 PM
One thing that just keeps giving is @vincentab.bsky.social's marginaleffects package (see marginaleffects.com). It's just a complete game changer, moving from "trying to discern what model coefficients tell you after careful recoding" to "you know what, just compare these slopes, thx"
Burlesque name Gini Coefficients
February 7, 2026 at 5:20 AM
Burlesque name Gini Coefficients
you repeat this process over and over again until the model can predict what is going on. but there is going to be a problem.
what we are really doing here is multiplying matrices with extremely small coefficients by other matrices with very small coefficients. the computer has finite precision.
what we are really doing here is multiplying matrices with extremely small coefficients by other matrices with very small coefficients. the computer has finite precision.
October 13, 2025 at 6:56 PM
you repeat this process over and over again until the model can predict what is going on. but there is going to be a problem.
what we are really doing here is multiplying matrices with extremely small coefficients by other matrices with very small coefficients. the computer has finite precision.
what we are really doing here is multiplying matrices with extremely small coefficients by other matrices with very small coefficients. the computer has finite precision.
which is the worse statistical sin, multiplying regression coefficients (mediation) or dividing regression coefficients (instrumental variables)
December 1, 2025 at 8:53 AM
which is the worse statistical sin, multiplying regression coefficients (mediation) or dividing regression coefficients (instrumental variables)
bluesky toxic coefficients:
September 17, 2026 at 4:40 AM
bluesky toxic coefficients:
Sociology: you can’t compare logit or probit coefficients across models
Psychology:
Psychology:
3. I then found out that the author mixed suicide RATES (per 100,000) with suicide RATE RATIOS either standardized to 1 or to 100. See my comments to the figure. I confirmed this by looking at several individual studies. Of course this doesn't make sense at all.
August 18, 2025 at 7:12 AM
Sociology: you can’t compare logit or probit coefficients across models
Psychology:
Psychology:
It never adds up.
Good morning! ☕️☕️
Good morning! ☕️☕️
April 7, 2024 at 10:40 AM
It never adds up.
Good morning! ☕️☕️
Good morning! ☕️☕️
You're very welcome @vincentab.bsky.social
P.S. Check out our preprint on an alternative to staring at coefficients: j-rohrer.github.io/marginal-psy...
P.S. Check out our preprint on an alternative to staring at coefficients: j-rohrer.github.io/marginal-psy...
September 17, 2025 at 1:32 PM
You're very welcome @vincentab.bsky.social
P.S. Check out our preprint on an alternative to staring at coefficients: j-rohrer.github.io/marginal-psy...
P.S. Check out our preprint on an alternative to staring at coefficients: j-rohrer.github.io/marginal-psy...
November 18, 2024 at 10:21 PM
"Here are my regression coefficients. I'm sorry the numbers on the slide are so small. I'll read them to you."
April 2, 2025 at 8:34 PM
"Here are my regression coefficients. I'm sorry the numbers on the slide are so small. I'll read them to you."
He’s got the best wind tunnel coefficients👍
September 10, 2026 at 3:27 PM
He’s got the best wind tunnel coefficients👍
Problem with this study is that they’re comparing positive and negative signed coefficients that are both not significantly different than zero. An honest study would’ve treated insignificant coefficients as zero.
I told authors this and they didn’t care as they wanted the headline.
I told authors this and they didn’t care as they wanted the headline.
I take no pleasure in reporting that social scientists' conclusions regarding the effect of immigration on support for the welfare state, based on the exact same data, are affected by their own political beliefs about immigration. Wild experiment by Nate Breznau et al. www.nber.org/papers/w33274
Ideological Bias in Estimates of the Impact of Immigration
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...
www.nber.org
August 19, 2025 at 6:23 AM
Problem with this study is that they’re comparing positive and negative signed coefficients that are both not significantly different than zero. An honest study would’ve treated insignificant coefficients as zero.
I told authors this and they didn’t care as they wanted the headline.
I told authors this and they didn’t care as they wanted the headline.
When the standard errors of all of your coefficients are missing.
September 27, 2024 at 6:11 PM
When the standard errors of all of your coefficients are missing.
how are you going to ban generally having coefficients for high order functions
September 13, 2026 at 1:02 AM
how are you going to ban generally having coefficients for high order functions
Alex Stubb on EU competitiveness - excellent
"Economists look at crude GDP... but I look at our Gini coefficients.. & whether it's justice, least corrupt, open, welfare, environment, open media, gender equality.. I don't buy the argument we're doing bad. I'm actually bloody happy"
MSC '26
"Economists look at crude GDP... but I look at our Gini coefficients.. & whether it's justice, least corrupt, open, welfare, environment, open media, gender equality.. I don't buy the argument we're doing bad. I'm actually bloody happy"
MSC '26
February 14, 2026 at 2:56 PM
Alex Stubb on EU competitiveness - excellent
"Economists look at crude GDP... but I look at our Gini coefficients.. & whether it's justice, least corrupt, open, welfare, environment, open media, gender equality.. I don't buy the argument we're doing bad. I'm actually bloody happy"
MSC '26
"Economists look at crude GDP... but I look at our Gini coefficients.. & whether it's justice, least corrupt, open, welfare, environment, open media, gender equality.. I don't buy the argument we're doing bad. I'm actually bloody happy"
MSC '26
The ideology of Economics fields.
Source: academic.oup.com/ej/article/1...
Source: academic.oup.com/ej/article/1...
December 17, 2024 at 6:36 AM
The ideology of Economics fields.
Source: academic.oup.com/ej/article/1...
Source: academic.oup.com/ej/article/1...
Listen, I'm gonna multiply these two coefficients. If you claim I'm doing a mediation analysis, that's on you.
August 21, 2025 at 9:03 AM
Listen, I'm gonna multiply these two coefficients. If you claim I'm doing a mediation analysis, that's on you.
we’re trying to pull like correlation coefficients from relationships which are caused by viral tiktok videos or boomer facebook memes, which are about as predictable as lightning strikes
June 22, 2025 at 10:58 PM
we’re trying to pull like correlation coefficients from relationships which are caused by viral tiktok videos or boomer facebook memes, which are about as predictable as lightning strikes
As far as I know, these are all the equations between binomial coefficients that don't follow from a few well-understood infinite patterns.
In 1997, de Weger conjectured there are no more!
deweger.net/papers/[20]d...
In 1997, de Weger conjectured there are no more!
deweger.net/papers/[20]d...
September 20, 2026 at 1:47 PM
As far as I know, these are all the equations between binomial coefficients that don't follow from a few well-understood infinite patterns.
In 1997, de Weger conjectured there are no more!
deweger.net/papers/[20]d...
In 1997, de Weger conjectured there are no more!
deweger.net/papers/[20]d...
Sinister bourgeois heat pumps with their high coefficients of performance.
February 10, 2026 at 7:49 AM
Sinister bourgeois heat pumps with their high coefficients of performance.
Agree with your assessment of buckwildery coefficients.
Studying Testing
ALT: Studying Testing
static.klipy.com
September 25, 2026 at 1:57 AM
Agree with your assessment of buckwildery coefficients.
My student Hannah Zadeh just completed an excellent dissertation on the inclusion of race coefficients and predictive stats more generally in medicine.
a main takeaway: AI tools lock in on these factors in part because **researchers have been doing the same thing**
www.proquest.com/openview/f22...
a main takeaway: AI tools lock in on these factors in part because **researchers have been doing the same thing**
www.proquest.com/openview/f22...
www.proquest.com
December 20, 2025 at 3:57 PM
My student Hannah Zadeh just completed an excellent dissertation on the inclusion of race coefficients and predictive stats more generally in medicine.
a main takeaway: AI tools lock in on these factors in part because **researchers have been doing the same thing**
www.proquest.com/openview/f22...
a main takeaway: AI tools lock in on these factors in part because **researchers have been doing the same thing**
www.proquest.com/openview/f22...
i control the coefficients of the convolution
August 6, 2026 at 4:38 AM
i control the coefficients of the convolution