I had my friends vote on which set of positive and negative words described me, then took the top 16 words and fed it to chatGPT to ask it what my astrological sign is.
It told me I'm an Aquarius. Then I asked-
"You're a professional astrologer. You're here to determine the top astrological sign that matches a set of personality traits, from most to least corresponding. The traits are: "[traits]" As a professional astrologer, what is the top matching astrological sign?"
That's good info, but I'm more concerned about the set of descriptors, how you selected them, etc.
I intuit that outcomes can be steered significantly by a biased set.