I’m unsure, and would appreciate others’ perspectives. I presented it as I did to avoid any misreading of percentages as implying that these were representative of all EAs in the world, as explained in my reply to Ryan. Perhaps that was overcautious, as we could clearly state that these are percentages of survey respondents rather than all EAs in the world, who’ll have somewhat different proportions. The absolute numbers of people reporting different answers are “count data”, and hard to misinterpret.
Do you think maybe this data would be more usefully presented as percentages of survey respondents rather than raw numbers?
I’m unsure, and would appreciate others’ perspectives. I presented it as I did to avoid any misreading of percentages as implying that these were representative of all EAs in the world, as explained in my reply to Ryan. Perhaps that was overcautious, as we could clearly state that these are percentages of survey respondents rather than all EAs in the world, who’ll have somewhat different proportions. The absolute numbers of people reporting different answers are “count data”, and hard to misinterpret.
Yeah just so long as you’re clear that these are proportions of your sample and not all EAs, it’s fine.