Ah yes. This must be the singular study that everyone gets these percentages from. But I didn't know it was by phone! Blech. Did they collect demographic information as well? I'd guess that the types of women that would answer these were probably all elderly, motherly women. It would be nice to have a more updated survey, perhaps one that tries to gain the participation of a younger crowd. But then, that would be a significant amount of money "wasted" in a rather meaningless study.
It was done by phone, and was rebalanced in various ways to correct for an over-representation of white females (for example) and an under-representation of black males (as another example). The sample was rebalanced to reflect census data for gender and ethnic makeup of the US population.
They both tried techniques of ignoring the excess in the over-sampled groups, and rebalancing the weights of the various groups. Both yielded about the same percentage representation.
I'd agree that doing anything by phone is likely to affect the weight of the distribution of participants (with, hypothetically in MBTI terms, Extraverts and Feelers tending to be more represented). I think it'd be tough to do better without doing door-to-door, census style canvasing (which would be hugely expensive, and even more time intensive than the actual census).
So, yes: what we have is imperfect and is affected by the means of data collection. I'd love to see better data, but that's not what we have, currently.
Still, even in Big Five terms:
wikipedia said:
A study of gender differences in 55 nations using the Big Five Inventory found that women tended to be somewhat higher than men in neuroticism, extraversion, agreeableness, and conscientiousness. The difference in neuroticism was the most prominent and consistent, with significant differences found in 49 of the 55 nations surveyed. Gender differences in personality traits are largest in prosperous, healthy, and more gender-egalitarian cultures. A plausible explanation for this is that acts by women in individualistic, egalitarian countries are more likely to be attributed to their personality, rather than being attributed to ascribed gender roles within collectivist, traditional countries.[74] Differences in the magnitude of sex differences between more or less developed world regions were due to differences between men, not women, in these respective regions. That is, men in highly developed world regions were less neurotic, extraverted, conscientious and agreeable compared to men in less developed world regions. Women, on the other hand tended not to differ in personality traits across regions.[75]
So that suggests by correlation to MBTI preferences (but, because that's a correlation of a correlation, doesn't prove much), that women tend to be more E, F, and J than men... and that, in turn, matches the data from the MBTI National Representative Sample. It's not definitive proof, by any means. Still, if you can point me at any better data, I'd be all ears.