Quentin André

I wish journalists reported more carefully on science. I really do. A short thread. The study cited at in the intro of this article is a mice study. Researchers randomize male to have access to a low/med/high amount of ethanol... 1/N www.theatlantic.com/family/2026/...

People are inundated with restrictive and contradictory health recommendations. Trust in public health is plummeting. So when you are writing about health and medicine, you have a unique responsibility to get things right. 11/N

I wish journalists reported more carefully on science. I really do. A short thread. The study cited at in the intro of this article is a mice study. Researchers randomize male to have access to a low/med/high amount of ethanol... 1/N www.theatlantic.com/family/2026/...

Finally, back to the sample size issue: The authors state Control and Low were topped up from "a separate experimental cohort." Why? They don't tell. How do they account for this in their analysis? They don't. 9/N

I wish journalists reported more carefully on science. I really do. A short thread. The study cited at in the intro of this article is a mice study. Researchers randomize male to have access to a low/med/high amount of ethanol... 1/N www.theatlantic.com/family/2026/...

But then they conduct statistical test of differences between groups on these scores! That's right: They tests for group differences *on variables that were constructed to maximize group differences*! No wonder they report p-values not even found in high-energy physics. 7/N

I wish journalists reported more carefully on science. I really do. A short thread. The study cited at in the intro of this article is a mice study. Researchers randomize male to have access to a low/med/high amount of ethanol... 1/N www.theatlantic.com/family/2026/...

Any statistican who has seen this before (insanely small p-values with small N) knows the most likely culprit: "Pseudoreplication", when non-independent observations are analyzed as if they were independent. Claude Opus 4.8 provides an exellent summary of the issue. 5/N

I wish journalists reported more carefully on science. I really do. A short thread. The study cited at in the intro of this article is a mice study. Researchers randomize male to have access to a low/med/high amount of ethanol... 1/N www.theatlantic.com/family/2026/...

This is expected with such low power: There are 8 sires in Low, 4 in Medium, 4 in High (yes, cells are imbalanced. I'll come back to this). Now, with such low power to detect the impact of the *manipulation*, you'd expect even less power to detect downstream effects. 3/N

I wish journalists reported more carefully on science. I really do. A short thread. The study cited at in the intro of this article is a mice study. Researchers randomize male to have access to a low/med/high amount of ethanol... 1/N www.theatlantic.com/family/2026/...

... have them sire offsprings, and measure the craniofacial features of the fetuses. Now, I don't know much about animal studies or biology, but I know about statistics and experimental design... and after spending five minutes on the paper, I could identify MANY issues. 2/N

HBS' latest filing in the Gino lawsuit contains a few new details: 1. A committee of 26 professors, admins, and fellows reviewed the allegations, and *unanimously concluded* that there was *clear and convincing evidence* of research misconduct.

She also refused to deny or admit that the file had been backdated by her or someone else. 4. She provided no evidence of sex-based discrimination. She only cites a private statement from her friend Gary Pisano... who later testified he couldn't identify any evidence of bias.

HBS' latest filing in the Gino lawsuit contains a few new details: 1. A committee of 26 professors, admins, and fellows reviewed the allegations, and *unanimously concluded* that there was *clear and convincing evidence* of research misconduct.

2. Contrary to Gino's argument that she was outlawyered and limited to two advisors, HBS says that she was, from the beginning, assisted by a team of attorneys "experienced in representing respondents in research misconduct cases".

This hasn't received much attention, but it looks like scammers also hacked into Francesca Gino's MTurk account, and asked MTurkers to buy her book Rebel Talent to boost the first-week sales of the book. Poor lady cannot catch a break. (links in comments)