I've been struggling to figure this one out for my whole life. It's quite clear that I seem to have strong aspects of both male and female in my personality.
This model is simplistic, and wrong. It's not sophisticated enough. A more useful model is this one, where masculinity and femininity are independent traits. An individual can therefore be either mostly masculine or mostly feminine; strong in both at once, or even not particularly strong in either. This diagram plots masculinity and femininity as two independent bars on a chart, but it would work equally well to plot them as two dimensions on a graph.
Before I go further, let me be clear that this model says nothing about the gender of the person you would most like to go to bed with. In other words, this model describes gender behaviour but not sexuality.
My results for this test were very interesting (at least to me). The results show that when it comes to masculinity, I am substantially above average across all categories. One way to interpret that result is to say my brain is more masculine than the average man's brain. When it comes to femininity, I am also above average in all categories (less so than for masculinity). One way to interpret that result is to say that my brain is also more feminine than the average woman's. Therefore I seem to score very highly for both masculinity and femininity. This seems to fit with the way I understand myself, inasmuch as I have been studying my own behaviour for almost four decades.
But once again, it's not as simple as that. For one thing, this test is not properly validated, although it's a clever and interesting diversion and the results are thought-provoking. In addition, I am sure people fluctuate in their performance and mood: I know that some days I feel very girly and other days I feel very manly. That might or might not influence my performance on the parts of the test.
Secondly, it makes no correction for IQ. I know my IQ, and it suggests I am likely to be intrinsically good at doing these sorts of tests. In other words, I am likely to have artificially higher results in both parts because of IQ, not gender alone. My final annoyance with the BBC test is that, although it displays the results for masculinity and femininity separately (which is great), on the last page it tries to boil that down to the linear scale above! It tells me my brain is 25% male.
Most recently, I've had my personality extensively and professionally assessed. Although the test gave results which I think fit very closely with my own views above, and seem to be very predictive of my reactions in some circumstances, the results made no mention of masculinity or femininity at all. I found that most refreshing. It fits with an even more fluid model of human behaviour, which I have tried to illustrate here.
What are these graphs counting, or showing? Nothing, except some hypothetical quality of behaviour. In themselves, the terms masculine and feminine are pretty meaningless: there is no absolute scale against which to judge either of them. The only way they can be defined is recursively: masculinity is what men are like. What are men like? They are masculine.
I think all of these models are attempting to make sense of an extremely complicated phenomenon. The subtleties of human behaviour and motivation are far too elusive to be nailed down in simple graphs, no matter how complex the test. Having done all these tests, I have come to a much more complete understanding (and acceptance) of myself as a whole person. I know that Vivienne isn't some alter ego of me: she is me. She cannot be suppressed, or got rid of, and she brings me a suite of skills and attitudes which have proven to be very helpful in my life.
But it's one thing to accept myself: what society thinks is a whole different ballgame.
Addendum: 28th April 2012.
Thanks to John, who posted below, I have found an insightful and interesting blog post by Jack Molay, which can be found here. It explores these same themes from a slightly different angle and in much greater depth.