I've read about studies [1] which shown that boys and girls under 1 year old already show preferences which we attribute with stereotypical gender roles (boys were more appealed to mechanical objects while girls were more interested in human faces (=>social interactions)).
[1] Since there are also studies which show that a nontrivial part of results in social science papers are not replicable, it has to be taken with a tiny grain of salt.
[1] Since there are also studies which show that a nontrivial part of results in social science papers are not replicable, it has to be taken with a tiny grain of salt.