that guy with the panda avi wrote a nice paper out it link.springer.com/article/10.1...

Any reasonably trained psychometrician would alert you to the fact that you have no outcome, you have no valid decision rule, you lack measurement equivalence and thus any justification for using your measures. I'd like to think they might also point out your moral failings...2/

You mean using half-ass psychometrics to cover their full-ass moral failing. I mean, the irony of going with the "we got to measure things well" argument just kills me. At some level, their instinct is right, but only if they really mean measure things well. 1/

Did they ever consider alternative approaches like simply providing more governmental support for the kids that they flagged instead of traumatizing the parents? I'm assuming all of the answers to these questions are no... 4/

Did they ever bother to do a study of the effect of ripping kids away from their birth parents, or did they just assume they were doing something positive? Did they ever consider that the disparate impact on Greenlanders was racist? 3/

Did they ever show that their concept of parental competence predicted anything of importance in the kids outcomes? Did they even bother to test the measurement equivalence of the tests given the different language and culture? 2/

Wow, just wow. There are so many questions...Did they ever define "parental competence" in a justifiable way or just wing it? Did they ever do a study showing that any part of their assessment could predict their definition of "parental competence"? 1/

It was a depressing if funny experience. We wouldn’t have noticed it if not for the amazingly high covid rate in the sample. Made us take a closer look and…

I agree the botpocalpyse is not yet upon us, but nanobrowser is scary easy to use. bsky.app/profile/rube...

As long as those bad actors can be detected...I do wonder whether "authenticity" (e.g., I'm a real human; I actually wrote my test answers) will be the most valued capital of the future.

That said, it might not take much. We had one person screw up a Covid survey by posing as 80 different people--he wanted the $20 per completed survey. All it would take is one or two talented people to render the entire system untrustworthy.

Oh quite the opposite. Amercans love their learning. Especially when we can "do our own research" thus making it more likely we'll be swayed by disinformation peddled by neerdowells from our country and abroad.

Pardon me while I down a few more Tylenol...

You mean, like Germany and many other countries do?