What do men exaggerate and women downplay?
Their computer skills. Eszter Hargittai and Steven Shafer put 51 women and 49 men through
internet-use skill test, and found that the women consistently rated their skill lower than comparably skilled men. The skills test gave participants a number of subjects to research, and allowed them to search for the information however they saw fit (with the researcher watching over their shoulder).
Hargittai and Shafer found no statistically significant difference between actual skill levels for men and women. However, "not one woman thought of herself as an `expert' user, and not one man thought of himself as a complete novice." The sex of the researcher administering the survey and observing the web use was not mentioned, and that seems like an oversight. The survey about web use was administered orally to "establish rapport." It would be interesting to see how results differed with less face-to-face interaction with researchers, or with an interviewer of a different sex. As was demonstrated by
research done by BMW, directions from a man and from a woman can be met with very different responses. A similar experiment, with male and female researchers and an incentive to perform well would likely yield different results.