Also, to bring together
, or fail to distinguish
, particularly in statistics
al work. Often happens because, while one group may tend towards having an attribute
which causes experimental result
A, and another group has a different attribute, causing experimental result B, it's the attribute, and not the group membership, that's causing the result. Group membership and attribute are being conflated.
For example (a daft example): more women than men have long hair. If I were to ask a group of men and a group of women whether they use elastic hairbands, I'd almost certainly find a statistically significant difference: more women than men would answer 'yes'. I could conclude that men don't like hair accessories, and that Company X should market them in masculine colour schemes and designs, to appeal to men. I'd be pretty stupid to do so: men are simply less likely to need to tie their hair back. I'd've been conflating gender and hair length.