Originally Posted by
Global_Hi_Flyer
If you're trying to determine response to advertising, then "self selection" is what you're doing. After all, those folks are choosing to buy product based on the ad.
By offering an incentive tied to the ad (discount, etc), then you tie the ad to consumer response. Otherwise, the response may just be chance.
You could actually do a clipboard corps airside and approach targets at random, asking if they recalled the display. That wouldn't be self-selection. While it would be difficult in most outdoor/display/transit environments where people are in constant flux, an airport would be an OK venue for that. You would only be testing for recall, opinion modification, and possibly intent to buy, though, not for actual follow-thorough.