K. Matersmore apparently does not understand “margin of error” for surveys.
They took a few questions (different than first) rated on scale 1 to 5? If I recall … maybe different. Definitely not the scale 1-100 they report. Goes up 1 point (in a hundred?) and they claim across the board victory. Of course maybe it isn’t 100… maybe just some made up metric entirely with no meaning at all.
These guys must have worked for the board of education. Sounds like the same math they use to grade final exams in American public schools. They start with the average they want and curve fit the results to match.