That being said a lot of those canned questions are designed to put you in a specifc mood for the rest of the study. It's not so much that the researcher wants your answer. They more want you to spend some time thinking about a certain stimuli before proceeding to the rest of the study.
I'd assert that they're not necessarily getting the mood they're after when dealing with anyone who has answered the question before.
Researchers are really really good at collecting all kinds of data, formatting it in specific ways, and then coming to an erroneous conclusion using it. I see this all the time on PubMed. This arrogance by researchers, assuming they know what data "means", is why the pollsters were all so surprised that Trump was elected president when many more intelligent people realized it was inevitable.
When you've answered whether or not you're going to kill the baby for the 50th time, you're not feeling guilty about your decision. You're just feeling irritated about being asked the same thing for the 50th time. If the researchers are asserting that thinking about that scenario will make you feel one way or another, they're going to come to erroneous conclusions from their research. Period.