This measures style, not authorship. Humans write slop and models can avoid it. Do not use it to make decisions about people (grading, hiring, misconduct).
This page needs JavaScript; it runs entirely in your browser.