If the average word is 4 syllables and the average sentence is 4 words, that gives a “grade level” of 33+. I imagine the problem with the ballot measures are extremely long sentences. Lists, conditions, and lists of conditions all make for very long sentences that can still not require any sort of advanced vocabulary. If the average sentence is 65 words and the average word is just 2 syllables, you still get a 33+ “grade level”.
So it has nothing to do with the content at all, just how many words are in sentences, and how many syllables are in those words. A twelve year old with a penchant for run-on sentences (and a thesaurus) could invent a “33rd-grade” sentence if they put their mind to it.
It’s actually dumber than if they had an AI grade it for readability.
That’s like saying the ability to have theoretical shoe sizes for giants makes practical shoe sizes useless. In reality, the hypothetical text you’re describing that breaks the scale isn’t something we need to worry about.
And yet, I’ve seen it touted as a useful, if approximate, metric since I was in secondary school last century (1990s).
Sometimes the “dumbest” stuff gets used just because it’s a translation to the “standard” number line. I don’t know which, if either, was the cause and which was the effect, but it kind of thinking strongly correlates with a “number go up” / infinite growth measure of success.
They explain what absurdly high “grades” mean earlier in the article.
I have to assume this is part of a standardized reading-level scale, not something they made up for the article.
Oh, I think they made it up. Most PhD’s “only” need 10-12 years of school after High School. 33rd Grade isnt a thing at all.
Long words and long sentences can get non-sensical “grade levels” out of https://en.wikipedia.org/wiki/Flesch–Kincaid_readability_tests#Flesch–Kincaid_grade_level
If the average word is 4 syllables and the average sentence is 4 words, that gives a “grade level” of 33+. I imagine the problem with the ballot measures are extremely long sentences. Lists, conditions, and lists of conditions all make for very long sentences that can still not require any sort of advanced vocabulary. If the average sentence is 65 words and the average word is just 2 syllables, you still get a 33+ “grade level”.
So it has nothing to do with the content at all, just how many words are in sentences, and how many syllables are in those words. A twelve year old with a penchant for run-on sentences (and a thesaurus) could invent a “33rd-grade” sentence if they put their mind to it.
It’s actually dumber than if they had an AI grade it for readability.
That’s like saying the ability to have theoretical shoe sizes for giants makes practical shoe sizes useless. In reality, the hypothetical text you’re describing that breaks the scale isn’t something we need to worry about.
And yet, I’ve seen it touted as a useful, if approximate, metric since I was in secondary school last century (1990s).
Sometimes the “dumbest” stuff gets used just because it’s a translation to the “standard” number line. I don’t know which, if either, was the cause and which was the effect, but it kind of thinking strongly correlates with a “number go up” / infinite growth measure of success.
Which should tell you something when we have ballots that go that far off the scale. It’s not an accident.