Experience is but one dimension of orientation, but as a dimension, I think it would be incomplete and therefore unfair of me to leave it out. I do not specialize in stats, survey design, or LGBTQ studies, but my first inclination would be to weight the orientation that someone identifies as, as the highest, and then go with their attraction, ... and oh dear, I seem, if we're to be looking at this more completely, to have gone straight from attraction to sex, and left out physical expressions of attraction that are not actually sex (daring hand-holding under the desk in class to the sort of heavy petting that's not sex because someone found a technicality) and also both short and long term partner-bonding.

I want to be as complete as possible so that ten years down the road, someone who comes across the survey will not have the opportunity to curse me for being imprecise and leaving stuff out. I want to play with the survey such that someone can answer one or two questions and satisfy the original reason (so how about those not-straight-female slash fans then?), then submit it and head off, or go through and answer a whole tl;dr set for let's-shake-the-data-and-see-what-else-falls-out purposes.