We all know the dangers of using oversimplified models in other contexts, but the same applies here and happens to be one of the largest generators of present conflict. It's a classic map-territory problem applied to people themselves and - whether the map makers know this or not - those who control the maps can sometimes also control the territory.
Yes, giving identical labels to very different phenomena is a big problem in this kind of survey-based approach.
On Maggie McNeill's blog, one post touched on the idea that women want sex to last longer. She pointed out that women do frequently say this, but what they mean is that they want foreplay to last longer.
Basically this just tells me that people lie on personality tests, or they interpret questions differently. Behaviorism is the only science in psychology. Question answering should be considered a behavior, and not a reflection of internal characteristics. In that light, you essentially have, among men mostly, a behavior, direct presentation of mating intentions, which has been consistently punished, and the opposite behavior, reflection of female mating desires, which has been inconsistently rewarded. The result is either duplicity or avoidance. It seems that the reverse phenomenon may occur for expectations later in the relationship, as evidenced by females initiating 90% of divorces among college-educated women.