those levels weren't what I was expecting.

https://nces.ed.gov/surveys/piaac/measure.asp?section=1&sub_...

I'm curious as to how I would score, I would definitely count myself as "literate" but I wonder how well I'd do on the level 4/5 tasks and if they cross over into more general memory, intelligence, and study habit metrics that even a normally "literate" person would not do well at.

Though given those descriptions I can't help thinking those would be great tests for AI. I'd love to see the proficiency scores for various models.

EDIT: Ok I just needed to scroll further, they have sample items in the last section up to level 4 and even at level 4 the question seemed trivial.

The most wordy one is the Q Drum article (which by the way Q drum is a real thing, kinda neat idea) and there's literally only two basic criticisms (flat land and expense) and if you had any idea what the life straw is you can probably construe what the similar criticism in the email is going to be without even looking.

Based on the scores and the proficiency description I assumed they were actually targeting some sort of normal distribution and levels 4/5 would be genuinely difficult explaining the scores. I'm now much more sad that the scores are so low.

At least I got a laugh at how they refer to each test item as "the stimulus" which has such a sterile/clinical flavor to it.