Got a survey from a Berkeley student group who are “researching how AI-powered tools can support educators like yourself by improving teaching efficiency and the learning experience,” and started filling it out just out of curiosity…
…but bailed on the survey when I reached this (mandatory) question. The correct answer for me, not present, is “I rarely use it •because• I am very familiar with it.”
I feel like there’s a classic social science and/or philosophy term for the way a question can circumscribe its possible answers, and thus skew our thinking. What is that? It escapes me. Whatever it’s called, this is a classic example.
You can see the students’ assumption / agenda plain as day: “The only barrier to AI utility is unfamiliarity! Therefore we can help these poor benighted educators by making them familiar with it!”
“People tried it and it sucks” is not a thinkable thought.
Important detail: it’s not clear that accuracy is a goal of the poll. The group is not actually conducting an academic research project at all; they are a student-run “strategy consulting organization.” Sort of like Model UN but for McKinsey, I take it. Given that, I suppose this malformed poll is functioning exactly as it’s supposed to function, and giving true-to-life experience.
“Push poll” and “framing” both popular in the replies (and spot on), but the winner in my book is…
…wait for it…
“prompt engineering”
@inthehands begging the question?
When a poll does it, we usually call that a push poll
@jenniferplusplus @inthehands I wonder if it's worth talking to them about it because they're students, or do we think they know what they're doing?
@jenniferplusplus
Yes, push poll is a relevant thought! The thing I can’t tell is whether the poll authors know they have an agenda, i.e. is the sucker here me or them?
@inthehands why_not_both.gif
@jenniferplusplus
Yeah, of course you’re correct: the best con artists believe their own con, or are incapable of actual belief altogether — and you’d think I’d know that after all these years of Musk and Trump and blockchain and crap
@jenniferplusplus I came here to say push-poll, so I'm glad to see you beat me to it.
@inthehands
@MattMerk @jenniferplusplus @inthehands Also “framing the discourse”.
@inthehands not exactly either of these things but somewhere in the ballpark of the concepts of the Overton window and push polling
@inthehands See “framing” à la George Lakoff. The Wikipedia article is super long so here is a pithier one:
https://commonslibrary.org/frame-the-debate-insights-from-dont-think-of-an-elephant/
@aarbrk
Yes, Lakoff for sure. Lakoff and…something out of Wittgenstein, probably?
@inthehands This is a really good example of poor survey design. Gawd, is nothing designed well??
@inthehands maybe "framing", in the George Lakoff sense, is the social science term you are thinking of? See posts at https://george-lakoff.com/tag/framing/ .
@wlf_warren Heh, “prompt engineering” is the best name yet for what this survey is doing!
@inthehands “leading question?” And generally “framing”?
@inthehands presuppositions?