But lets do the math: Each heterosexual act involves one male and one female. If it's their first encounter with one-another, then
Clearly, the two totals are equal. We also know that the US has about the same number of males as females in the 20-59 age range. Calculating the average number of partners for males and females, we have the same numerators and same denominators, so there's no way around the fact that the average number of heterosexual partners for males must equal that for females.
So, that's "The Frakking Paradox." Explanations I've considered but ruled out:
Guys are frakking lots of females during overseas vacations (but then they'd pretty much be dead, according to current cinema, and not be able to participate in the survey)
Guys are frakking lots of female aliens who are visiting terrestrially (but wouldn't the male aliens come on a visit, too?)
Guys are frakking lots of female aliens during an abduction (being probed would probably interfere, but then again, who knows?)
US females change their clothes, cosmetics, hair, and even have surgery; but US males are so stupid, they think she's a new partner, while the females realizes it's the same old slob of a guy in the same pair of jeans as always. (Nah, if he's that much of a jerk, she's liable to take him out of the running permanently, if you catch my jist, which would then skew the results in the wrong direction.)
There's something that males and females can do together that the male counts as having had a sex partner, but not the female. But if this were possible, then slashdot readers wouldn't be such a sorry lot. (Note that the wording of the original poll seems to rule out various asymmetries you might be getting your hopes up for.)
They measured the "median", not the "average". But this only helps if there are many, many more females than males who have a wildly high number of partners. Even considering prostitution, you'd have to have almost all males going to about four different prostitutes each, and very few males going to many of them.
Guys are frakking lots of females who are outside the 20-59 age range. This seems impossible given the rest of the data (age of first sexual encounter; median number of partners for 50+ females), and the relative rarity of google hits on "GMILF".
Lots of people are fibbing big-time. But the study authors went to all sorts of trouble creating an environment that's supposed to avoid this sort of trouble ("Audio Computer-Assisted Self-Interview... to allow respondents to answer questions in complete privacy").
Lots of people are self-deluding. But we're talking an almost 2-to-1 discrepancy here...