3 ms·
This is an appropriate argument basically any time somebody claims that [insert group here] [insert behavior of an individual here]. Groups of people do not beh
by nevinera 15y ago
This is an appropriate argument basically any time somebody claims that [insert group here] [insert behavior of an individual here]. Groups of people do not behave uniformly; when they behave in statistically significant ways, it is not for uniform reasons.
It's difficult to reason abstractly about groups as they actually are, so we abstract that group into a uniform body - that way we can make abstract claims about their behavior without the benefit of an education in applied statistics and research methodology. It's a useful tool for intuiting behavior, but try to avoid making non-statistical claims about the members of a group of people; you will almost never be correct.
- da_dude4242 15y agoI agree with you but in how many other cases are stats so clear? It's not like we're talking about slight margins of significants. ~80% of women have children and most of them do in what would peak years of career development. It doesn't explain why the female involvement for CS is so pathetically low compared to other traditionally male careers but I think it explains the values with which many woman are making lifestyle decisions.