All right. I'll bite.
I was constructive, Peter. I said he's in the wrong place. And that's far better advice than the pedantic bullshit you spewed about question "bias". Check your usage of that word, by the way. You're using it in a simple pejorative sense, and that's completely
inappropriate in the context of social research. Bias refers more to what I was talking about regarding sampling, which makes your rhetorical concerns pretty trivial by comparison.
Furthermore, it's perfectly acceptable to hide a hypothesis in survey questions. The trick is to make sure that it stays hidden. And it wouldn't constitute a form of confirmation bias if the sample is representative.
Sorry to be harsh, but this is a lazy approach to survey research, comparable to a journalist interviewing someone over an instant messenger program. You'll have much better luck if you use analytical methods, such as comparing the skin tones of models in magazine
advertisements. Jean Kilborne's
Killing Us Softly
lecture series is a must-see if you're talking about the way media influences our perception of beauty.
Sadly, the resources necessary for proper survey research are pretty much out of reach for undergraduates. Stick with the analytical stuff.