Crémieux Profile picture
Oct 4, 2024 • 10 tweets • 4 min read • Read on X
There are people who desperately want this to be untrue🧵

One example of this came up earlier this year, when a "Professor of Public Policy and Governance" accused other people of being ignorant about SAT scores because, he alleged, high schools predicted college grades better.Image
The thread in question was, ironically, full of irrelevant points that seemed intended to mislead, accompanied by very obvious statistical errors.

For example, one post in it received a Community Note for conditioning on a collider. Image
But let's ignore the obvious things. I want to focus on this one: the idea that high schools explain more of student achievement than SATs

The evidence for this? The increase in R^2 going from a model without to a model with high school fixed effects

This interpretation is bad. Image
The R^2 of the overall model did not increase because high schools are more important determinants of student achievement. This result cannot be interpreted to mean that your zip code is more important than your gumption and effort in school.

If we open the report, we see this:Image
Students from elite high schools and from disadvantaged ones receive similar results when it comes to SATs predicting achievement. If high schools really explained a lot, this wouldn't be the case.

What we're seeing is a case where R^2 was misinterpreted.
The reason the model R^2 blew up was because there's a fixed effect for every high school mentioned in this national-level dataset

That means that all the little differences between high schools are controlled—a lot of variation!—so the model is overfit, explaining the high R^2
This professor should've known better for many reasons.

For example, we know there's more variation between classrooms than between school districts when it comes to student achievement.

As another example, we know that achievement gaps exist along the whole continuum of school and district quality.

If the issue was really zip codes, high schools, and so on, this shouldn't be the case.

The other thing this professor should've known is that high school is biased! GPAs are biased too!

The bias in GPAs has actually been exploited: elite high schools inflate grades and don't report class ranks, so students appear better than they are. Image
But you know what isn't a biased tool for admissions? Just one thing: test scores.

Want to learn more? Here are some sources:

x.com/cremieuxrecuei…

cremieux.xyz/p/what-happens…

cremieux.xyz/p/bias-in-admi…

• • •

Missing some Tweet in this thread? You can try to force a refresh
 

Keep Current with Crémieux

Crémieux Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @cremieuxrecueil

Apr 3
Timeline's feeling down. Thread of good news.

At every age, the incidence of dementia is down. As a society, people are no longer suffering dementia nearly as often! Image
The world over, child mortality is way down. It's unusual for parents to experience the death of a child these days, where even a century ago, it was the global norm. Image
Each year, novel gene therapies are approved.

The number of gene therapies in the pipeline is also rapidly increasing. There is tons of progress to be made here, and the main issue is regulatory.

We have lots of low-hanging fruit in curing disease! Image
Read 13 tweets
Apr 3
Which is more impressive for China, its

GDP Per Capita, or its

Actual Individual Consumption Per Capita (AIC is a superior measure of how rich its people actually are. It's based on how much they really consume)

Answer is below.
Here's the GDP Per Capita data. Image
And here's the AIC data. Image
Read 5 tweets
Apr 3
There's a common type of misunderstanding that sounds like this:

"If taller people tend to be more educated, and women tend to be shorter than men, how do you explain women tending to be more educated?"

The issue has to do with intercepts. Consider this plot: Image
You can see that, among Whites, women tend to be shorter than men, and they tend to have lower earnings.

But at the same time, to similar degrees in both sexes, taller people tend to have higher earnings.

Perplexed? You shouldn't be.
The fact is that there's more to this that differentiates men and women than height, so the intercept for women is shifted down, even though the slopes of the height * income relationship are fairly comparable.

You could also ask the same thing about race.
Read 8 tweets
Apr 2
Debate about the value of essays in college admissions missed a key point:

Essays are biased, so should not be used.

Here's an example: High-income people know 'what to write' to look good to raters, so they outperform on essays relative to their other qualifications. Image
This shows up by race, too, and that's why admissions departments use essays to infer race for the express purpose of discriminating.

Write that you're Black; that you grew up as a poor immigrant; that you're gay or a cripple.

They love that!

The reason essays do not have a role to play in the admissions process is because they're biased. It's plain, it's simple, it doesn't need to be discussed any further.

And here's some good policy: Use tools that are not biased or lose public funding.
Read 4 tweets
Apr 2
A new meta-analysis and systematic review on the effects of social media abstinence interventions on mental health has been published.

First result: No effect on positive affect: Image
Second result: No effect on negative affect. Image
Third result: No effect on life satisfaction in general. Image
Read 6 tweets
Apr 2
Happy Autism Awareness Day! I think too many people are 'aware' of autism.

Have you ever met someone who claims to be autistic, but they've never been diagnosed?

Self-reported autism spectrum disorder (ASD) is practically uncorrelated with real, clinician-diagnosed autism🧵 Image
Sort self-reporters into those with high and low ASD scores, and you get the bars on the left. The "high-trait" self-reporters look like people with diagnosed autism (ASD column).

But they're more socially anxious (middle) and avoidant (right). Image
So far, the means of distinguishing diagnosed from self-reported autistics have been crude.

To get a more nuanced understanding of their differences, we have to look at behavior.

For that, we'll start with the social control task.
Read 14 tweets

Did Thread Reader help you today?

Support us! We are indie developers!


This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Don't want to be a Premium member but still want to support us?

Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal

Or Donate anonymously using crypto!

Ethereum

0xfe58350B80634f60Fa6Dc149a72b4DFbc17D341E copy

Bitcoin

3ATGMxNzCUFzxpMCHL5sWSt4DVtS8UqXpi copy

Thank you for your support!

Follow Us!

:(