3 ms·
> ...a full 95 percent of respondents were male Well, that’s exactly the problem I see with surveys in general: people read the above statement and automatical
by ManuelKiessling 6y ago
> ...a full 95 percent of respondents were male
Well, that’s exactly the problem I see with surveys in general: people read the above statement and automatically parse it as „95 percent of GitHub users are male“.
No, they aren’t. Or maybe they are. We don’t know. What we know is the percentage of respondents who identified as male. We know nothing - not. one. thing. - about how many % of GitHub users are male. What if 80% of female users simply didn’t bother to respond to the survey?
- AnthonyMouse 6y ago> We don’t know. What we know is the percentage of respondents who identified as male. We know nothing - not. one. thing. - about how many % of GitHub users are male. Ridiculous. We know the survey has error bars. Maybe they're large. But it's still information. > What if 80% of female users simply didn’t bother to respond to the survey? Then Github users would still be at least 84% male.