A recent international study about blood donors shows why asking direct questions of people is not always a good basis for decision making.
International guidelines prohibit monetary incentives for blood donation, based on concerns that any form of incentive may actually reduce motivation, at the same time as it puts the blood supply at risk from those motivated only by the cash rewards.
Extrinsic rewards (i.e. cash) can sometimes interfere with intrinsic rewards (feeling good about yourself).
The surveys showed one result
These policies have had support from surveys that ask people if they would donate blood based on receiving cash rewards.
"Policies that prohibit monetary incentives have been based primarily on population surveys in which respondents were asked if they would donate blood if offered a cash reward, and the majority nixed that idea. But Lacetera said responses were based on a hypothetical premise, not on studies of actual potential donors who were offered an incentive."
Nico Lacetera quoted in Globe and Mail
The real world experiments showed another
Actual experiments showed very different results. Donations increased for all sorts of things, like a day off work (Italy), a $5 lottery ticket (Switzerland), a $10 gift card (USA.)
What it all means: be careful with hypotheticals
We need to be careful when asking people direct questions about hypothetical situations. Human beings are great at coming up with answers, but our own ideas about our behavior are not always accurate. The futher removed we are from the real world situation the more this is true.
Running small experiments is a much better approach, whenever possible. In digital marketing and in things like loyalty programs, it is pretty easy to do this, and marketers are learning a lot from testing approaches.
But what if you can't run an experiment?
You should still try to avoid just asking the direct question. As a researcher, what I try to do is set up tasks that are similar, ask indirect questions, observe actual behavior, and otherwise try to overcome this problem.
Topics that are most likely to give you problems are direct questions involving motivation, hypothetical situations, and any topic where there are strong social beliefs or taboos, or where people might desire to paint themselves in a positive light. (A great many things, in fact!)
Here's a more detailed discussion from researcher Nicola Lacetera written in 2008
Here's the source of the scientific article in Science: Economic Rewards to Motivate Blood Donations published in 2013