Loading…
4
That stats prof who said 'show me the question wording' broke my trust in every poll graphic
Last week I shared a headline chart in a group chat showing that 68% of people support something, and my old stats prof (who never lets anything go) asked me one thing: what was the exact question? I looked it up and the actual survey had two versions, one worded as a yes/no and one as a 1 to 10 scale, and they mashed the top two boxes together to get that 68%. Only 41% picked the strongest answer when I dug into the raw tables. So now I always click through to the question text and the sample size before I share anything, and if I can't find them, I skip it. It takes like two minutes and half the graphics I see fall apart right there. How deep do you all go before calling a poll trustworthy?
0 comments
Log in to join the discussion
Log In0 Comments
No comments yet
Be the first to share your thoughts on this discussion.