Khuram's reference to what I call the "gap technique" is spot on. Few people will rate something 10/10 or A+. This gives you the opportunity to ask them how to achieve that highest rating. Once you do that: "Shut up and listen!" :) -- One of the worst mistakes is to get a participant talking (and many people talk way too much) but then to cut them off even though they're giving you substantive information (obviously, if it's drivel, then you do want to move on). You also have to assess those you survey: some are founts of information and insight while others have nothing to say. All are not created equal, as it were.

In all surveys, whether in person or on-line, I recommend open-ended (i.e., unprompted) questions (where possible) followed by close-ended (i.e., prompted answers) questions. In other words: What is your favorite brand of cookie (with no list specified)? Perhaps they say Oreos. Later on, after they've forgotten that, you have a list: Which is your favorite cookie brand: a) Little Debbie; b) Mrs. Fields; c) Oreo; d) Duncan Hines? Now if those two don't match up, how valuable are these answers? So that gives you the ability to test for validity. Once you present the prompted answers, you've poisoned the well, which is why they have to come later.

Hope this helps and should you have any questions, I would welcome the opportunity to discuss with you.

Best,
Steve