4 ms·
That thread also misinterprets the study. The question asked is multiple-choice, so you can't just add up the numbers the way he does (what's weird is that he a
by notJim 7y ago
That thread also misinterprets the study. The question asked is multiple-choice, so you can't just add up the numbers the way he does (what's weird is that he acknowledges this, and then adds up the numbers anyway…) Depending on how you imagine the percentages to overlap, you can kind of have it say whatever you want.
- iudqnolq 7y agoNo, he adds the numbers up to accurately provide a maximum value that's low enough that his point is proved regardless of how far below it the real value is. The survey design, reduced to a trivial example: Select all that apply - A: 100% select - B: 50% select - C: 12% select Suppose there were 100 people. There is no way to imagine overlap that leads to more than 12 people selecting C, as you claim.