While writers like Donald Symons and
David Buss have attempted to universalize their findings, and even purported to find
cross-cultural support for theories such as "sex is something that women have and men want", their research suffers from a multitude of both philosophical and empirical weaknesses. For instance, their samples neglect to include
matrilineal or
matrilocal societies, where sex differences in behavior are typically either minimized or reversed.
[50][51][52][53] Generally speaking, most samples which indicate the finding that men are more sexually motivated than women are from societies which are historically highly patriarchal, and hold (or held until recently) a highly negative attitude towards premarital sex, especially in women. Such societies tend to be large, populous states as opposed to smaller-scale
foraging and horticultural societies, which are usually more liberal towards sexuality.
[54][55] These studies almost always rely on quantitative self-report surveys, which are plagued with problems such as
social desirability bias (maximized for women in cultures which have traditionally valued
virginity), sampling issues and (in non-western cultures specifically) resistance or hostility to reveal personal information to outsiders. Cultural anthropologists working with
Boasian methodology with an emphasis on
participant observation and
non-judgmental,
emic evaluation of other societies have long warned of the dangers of reliance on quantitative methodology such as surveys (or any research method using a primarily etic perspective) for adequately capturing the internal beliefs and behaviors of other cultures.
[56][57][58][59]