Question Wording

Well, if our goal is to see if media effects politics directly, public perception of the effect media has should at least be something worth looking at. As I said above, I'm really not at all adamant about having the fifth question I wrote included. Feel free to delete if a better question comes up (we are supposed to have 5, no?).

As for "prefer not to answer" - it won't skew data if this response isn't included. If somebody chooses not to answer, you can just enter a null response when compiling data. I've personally only rarely seen "prefer not to answer" on a survey. What's more, I don't really think asking somebody if they voted is terribly personal - at least hardly more personal than asking them how often they use media. All that said, it's a pretty minor issue; including "prefer not to answer" doesn't really bother me. If "prefer not to answer" is an option for question 4, though, somebody will have to edit it "prefer not to answer" for other questions as well.

Finally, I like Kristen's formulation of the 4th question. It seems the least troublesome.